Background: A novel coronavirus caused an outbreak of acute infectious pneumonia are spreading over the globe. However, studies predicting prognosis are limited. We predicted outcomes of patients with coronavirus disease 2019 (COVID-19) using the neutrophil-to-lymphocyte ratio (NLR) on admission.
Methods: We retrospectively analyzed the characteristics of COVID-19 patients diagnosed from February 6 to March 1. The outcomes, including the occurrence of in-hospital mortality, acute kidney injury (AKI), and endotracheal intubation (ETI), were recorded. The relationships of neutrophils, lymphocytes, C-reactive protein, lactate dehydrogenase, and NLR with outcomes were assessed using multivariate regression model. P-values for trends across quartiles of NLR was examined.
Results: A total of 182 patients were included. 37 (20.3%) patients died during the hospitalization, 41 (22.5%) developed AKI, and 36 (19.8%) received ETI. The NLR had a superior predictive performance than others. Using an NLR cutoff of 11.4, the area under the curves (AUC) were 0.766 for in-hospital mortality, 0.755 for AKI, and 0.733 for ETI. In multivariate analysis, NLR >11.4 was further identified as an independent prognostic factor. Following stratification with quartiles of NLR, a positive trend between the increasing quartiles of NLR and the three outcomes were observed (p-values for trends across quartiles were 0.043, <0.001, and 0.041, respectively). The multivariate adjusted odds ratio (OR) in the highest quartile vs. the lowest quartile were 5.738 for mortality, 25.307 for AKI, and 5.136 for ETI.
Conclusions: Increasing NLR obtained on admission is a powerful predictor for inpatient mortality, AKI, and ETI in COVID-19 patients.