Early Neonatal Bacterial Infection (INBP) refers to an infection occurring within the first 72 hours after birth [1] . It is a neonatal infection transmitted from the mother to her newborn. It remains a leading preventable cause of neonatal morbidity and mortality worldwide and is a challenge for the clinician [2] . The clinical signs of these INBPs are often subtle and nonspecific in the newborn [3] . According to the World Health Organization (WHO), neonatal infections are the third leading cause of death, after preterm births and complications of childbirth (birth asphyxia) in developing countries [4] . Their incidence is higher in developing countries than in industrialized countries [2] .

Blood culture, which is the gold standard, is still not available. Apart from the fact that it could be compromised due to inadequate blood volume sampling in newborns, bacteriological confirmation is also lacking due to the great difficulty of isolating the germ in this category of infection [5] [6] . In developing countries, newborns suspected of infection are treated with empirical antibiotics. These antibiotics promote bacterial resistance to antibiotics, alter the gut microbiota, and increase the costs of care [7] . Some newborns may be completely asymptomatic in the early stages of infection. Therefore, such situations may lead to a delay in the treatment of infected newborns or treatment of uninfected cases, which is not without complications [8] . It is therefore necessary to improve diagnosis to better identify newborns requiring antibiotic therapy from those who do not [9] .

In the absence of methods for detecting the bacterial pathogen, the diagnosis of INBP is based on clinical signs and an increase in the concentrations of biomarkers such as Procalcitonin (PCT) [10] . Several publications note the fact that PCT is more specific for bacterial infections and can help distinguish bacterial infections from viral diseases. In addition, it has favorable kinetics for use as a clinical marker because it increases rapidly within 6 to 12 hours following stimulation [11] . According to studies, PCT would be an early marker of neonatal infection. It must be said that most of the studies that have demonstrated this have been carried out in developed countries [12] . In Sub-Saharan Africa (SSA), similar studies are almost non-existent. In the Democratic Republic of the Congo (DRC), to our knowledge, there are no studies that have addressed this topic. In Kisangani, one of the main cities of the DRC, a study evaluated the interest of CRP in the diagnosis of bacterial infection [13] , but no study has been carried out on PCT, hence the need to conduct this study whose aim is to evaluate the interest of PCT in the diagnosis of INBP in Kisangani.

A total of 123 newborns were divided into 3 groups, respectively of 5 certain infections, 81 probable infections and 37 no infections.

The median PCT in the definite infection group was higher 5.7 [3.1 - 9.4] than in the probable infection group 2.8 [1.0 - 19.1] with a significant difference ( Table 1 ). The threshold for PCT positivity was 2.97 ng/ml with a sensitivity and specificity of 80.0 % and 67.8 %, respectively ( Table 2 ). The ROC curve of PCT shows that the Area under the Curve (AUC) is 0.74 ( Figure 1 ).

The blood culture performed confirmed the infection of five newborns (4.1%): two had positive blood cultures for Enterobacter spp., two others for Staphylococcus spp. and one for Serratia spp.

¹Median [EI], n (%); ²Kruskal-Wallis test; ³Fisher’s exact test.

CI = 95% Confidence Interval.

INBP remains a concern for clinicians in low-income countries, especially in Sub-Saharan Africa, because of its difficult diagnosis and non-specific symptoms. This study aimed to assess the value of serum PCT in Kisangani hospitals. Indeed, PCT has been reported as a sensitive marker for early diagnosis. After injection of endotoxin, its concentration increases and is detectable after four hours and then reaches a peak after six to twelve hours [17] . In our study, the median PCT of the certain infection group was significantly higher than that of the probable infection group (5.7 ng/ml versus 2.8 ng/ml). PCT concentrations were found to be variable, although most values were between 0.4 and 6.6 ng/ml. These concentrations vary according to the studies, depending on the gestational age and the mode of delivery [3] .

In our study, the optimal cutoff value was 2.98 ng/ml with a sensitivity and specificity of 80.0 % and 67.80%, respectively. This result is similar to that found in Cameroon by Alima Yanda et al. [12] . PCT has a high sensitivity in bacterial infections, which would make it an ideal biomarker for INBP. The PCT threshold varies among studies depending on factors such as the incidence of early neonatal infection, the endpoints, and the study population [17] . PCT can be used as an important index for diagnosis and treatment evaluation in neonates with systemic bacterial infection [16] . Our result also corroborates that found by Maamouri et al. in Iran with high sensitivity and specificity [18] . Our study showed that the area under the curve AUC was 0.74, which demonstrates a good ability of the test to discriminate infected and sick neonates from those who are not infected. In a recent meta-analysis, PCT had a good discriminatory value for neonatal sepsis in low-income countries [19] . It is a reliable approach that can improve the quality of care and enable rational use of antibiotics in settings where access to reference diagnosis is limited.

PCT showed good performance in contributing to the diagnosis of INBP in Kisangani, a resource-limited setting. It was positive in all 5 cases of definite infections confirmed by blood culture. In resource-limited countries where blood culture is still not available, it can be used as a reliable marker for early identification of infected newborns.

We were unable to perform serial dosing. Also, the results were not correlated with gestational age and birth weight. Therefore, case-control studies on a large cohort of newborns should be conducted.

The study did not receive external funding.

The design of the protocol was provided by HMV, and the clinical examination of newborns suspected of INBP and the data collection were provided by the following authors: HMV, VMK, and DGM. The correction of the study protocol and the manuscript were provided by DFS, ETK and JPAO. All authors read and approved the final version of the manuscript.