BACKGROUND: Neonatal sepsis remains to be an important cause of neonatal morbidity and mortality and its diagnosis is difficult due to non-specific signs and symptoms that may mimic other infectious conditions. Blood culture, the gold standard in the diagnosis of sepsis, is limited by it being time-consuming and with high probability of false negative results.

OBJECTIVE: To investigate the usefulness of the NLR as a predictor in the diagnosis of neonatal sepsis and early - onset neonatal sepsis (EOS).

METHODS: Relevant publications from 2009 to 2019 that fulfilled the inclusion criteria were identified through electronic database search. Studies were analyzed and a meta- analysis was performed. The effect of NLR was calculated as a predictive factor for EOS. \

RESULTS: Four observational studies were included with a total of 392 patients. Two studies were analyzed for EOS which included 242 patients. There is significant association between NLR and neonatal sepsis. The sensitivity and specificity of NLR to predict sepsis were 84.5% and 91%. The sensitivity and specificity of NLR to predict EOS were 71% and 66%.

CONCLUSION: NLR is an acceptable tool in predicting neonatal sepsis and EOS but its usefulness is limited due to the presence of bias and heterogeneity in the studies included.

RECOMMENDATIONS: Further studies, preferably local studies, to investigate and validate the usefulness of the NLR as a predictor of neonatal sepsis and EOS is recommended.