Admission Neutrophil-lymphocyte Ratio for Prediction of Severity and Mortality in COVID Patients: A Systematic Review and Meta-analysis

ABSTRACT

Aims and objectives: Multiple studies have suggested that neutrophil-lymphocyte ratio (NLR) derived from differential white cell count might be a useful marker for COVID-related disease severity and mortality. We conducted a systematic review and meta-analysis and investigated if the same can be predicted with on-admission NLR values and also evaluated the prognostic significance of NLR on disease outcomes in patients with COVID-19. Materials and methods: We searched PubMed, EMBASE, MEDLINE, and SCOPUS databases for published articles in peer-reviewed journals from 01 March 2020 and 01 March 2021. Meta-analysis was performed to determine the pooled standardized mean difference (SMD) for the mean values of NLR. A random-effects meta-regression was performed for the following potential confounders age, gender, and comorbidities. Results: After study screening, systematic review included 68 studies comprising 15,818 patients in total, 2260 with severe disease and 1198 patients with mortality outcomes. The meta-analysis showed significant difference in mean NLR between severe and non-severe patients {SMD was 2.88 (95% CI 2.32 to 3.44)} and between survivors and non survivors {SMD was 7.89 (95% CI 3.37 to 12.42)}. Both outcomes were heterogeneous (Q = 1912.85, P < 0.0001, tau2 = 3.14, I2 = 98.35% and Q = 5898.15, p < 0.0001, tau2 = 116.65, I2 = 99.92% for severity and mortality, respectively). Meta-regression analysis showed that the association between NLR values on admission and severity in COVID-19 patients was not influenced by age (p = 0.893), cardiovascular diseases (p = 0.259), diabetes mellitus (p = 0.545), or hypertension (p = 0.104). Conclusion: On admission, NLR predicts both severity and mortality in COVID-19 patients and is not affected by age or comorbidities. Further high-quality studies are needed to confirm these findings. Results2 After study screening, systematic review included 68 studies comprising 15,818 patients in total, 2260 with severe disease and 1198 patients with mortality outcomes. A summary receiver operating characteristic (SROC) curve to determine a pooled estimate of the prognostic accuracy of NLR for severity showed that the pooled sensitivity, specificity and AUC were 80.2% (95% CI 74.0-85.2%), 75.8% (95% CI 71.3-79.9%), and 0.833, respectively, with the pooled diagnostic odds ratio of 13.63 (95% CI 9.71-19.02). According to the Q-test, the true outcomes appeared to be heterogeneous for both severity and mortality (Q = 1912.85, P < 0.0001, tau2 = 3.14, I2 = 98.35% and Q = 5898.15, p < 0.0001, tau2 = 116.65, I2 = 99.92%, respectively). The same was done for NLR and mortality showed the pooled sensitivity, specificity, and diagnostic odds ratio were 78.8% (95% CI 73.5-83.2%), 73.0% (95% CI 68.4-77.1%), and 11.483 (95% CI 7.814-16.875), respectively, with AUC of 0.820. Meta-regression analysis showed that the association between NLR values on admission and severity in COVID-19 patients was not influenced by age (p = 0.893), cardiovascular diseases (p = 0.259), diabetes mellitus (p = 0.545), or hypertension (p = 0.104).