Systemic infection is a medical emergency requiring early diagnosis and prompt management for the clinician. Delayed treatment increases the mortality rate [10, 14]. We conducted a cross-sectional study in 430 apparently healthy newborns in 24 hospitals in the city of Lubumbashi. The objective of this study was to assess the diagnostic performance of NLR in predicting systemic infection. The salient results showed that NLR better predicted systemic infection, relative to neutrophil, lymphocyte and platelet count.
The mean NLR values of patients with systemic infection were 4.3 ± 1.2, which is significantly higher than patients without systemic infection (1.9 ± 0.6). Our data are consistent with observations in the literature, including the work of Terradas et al.  and Zahorec . It is generally believed that NLR > 3.0 could symbolize positive blood culture results.
The critical value of NLR for the diagnosis of systemic infection was set at 3.0, its sensitivity 97.6% and its specificity 89.8%. Additionally, NLR is a good indicator of systemic infection with a subsurface curve of 0.885 (95% CI: 0.816–0.959). It is therefore necessary in poorly equipped settings where blood culture is not feasible, to be satisfied with NLR to enable a systemic infection to be diagnosed earlier.
Neutrophil count predicted systemic infection less, but more better than lymphocyte and platelet count. The sensitivity and specificity of neurophil count in predicting systemic infection was 98.2% and 88.5% and its area under the ROC curve was 0.738 (95% CI: 0.662–0.814). Our results relate to the literature, including the results of Charles et al , Brodska et al.  and Clec’h et al. , but unlike those of Dandona et al. , who believes that there is no significant difference in the prediction of systemic infection between NLR and neutrophil count. For Dandona et al, RNL is more related to the neutrophil level. An elevation in neutrophils should be linearly related to an elevation in NLR because they are closely related. This elevation of neutrophils often observed in systemic infection is due to chemotaxis due to the presence of bacteria in the body during the release of endotoxin as a pro-inflammatory cytokine . On the other hand, Surbatovic et al. had neutrophil and lymphocyte count predicted systemic infection better compared to NLR . The observed difference could be due to the divergence in methods and sample size between the two studies.
There were some limitations to this research. First of all, this is a cross-sectional study, the outcome of which was affected by other factors. We cannot know if other primary diseases, such as neutropenia, would influence the NLR value. Second, the rapid change in the condition of the newborn after birth can cause the NLR to change rapidly, which also affects the accuracy of statistical data.