Background: Since December 2019, coronavirus 2019 (COVID-19) has spread worldwide. Identifying poor prognostic factors is helpful for risk stratification. In this meta-analysis, we investigated the association between severe COVID-19 and a change in white blood cell (WBC) count, an elevation of C-reactive protein (CRP), and fever. Moreover, we aimed to evaluate the diagnostic accuracy of leukocytosis and an elevation of CRP.
Methods: We performed a systematic search of PubMed, EMBASE, Scopus, and the Cochrane Library through April 20th, 2020. The odds ratios (ORs) and 95% confidence intervals (CIs) were calculated. A sensitivity analysis was conducted according to the study size (>200 or <200) and median age (>55 or <55). Meta-regression analyses were conducted to examine possible sources of heterogeneity. We calculated the diagnostic accuracy of leukocytosis and CRP.
Results: Eighteen studies with 3278 patients were selected. Fever, leukocytosis, and elevated CRP were associated with poor outcomes (OR (95% CI) 1.63 (1.06-2.51), 4.51 (2.53-8.04), and 11.97 (4.97-28.8), respectively). Leukopenia was associated with a better prognosis (OR 0.56, 95% CI 0.40-0.78). Sensitivity analyses showed similar tendencies. Meta-regression analysis for leukocytosis indicated that age, dyspnea, and hypertension contributed to heterogeneity. The pooled area under the leukocytosis and CRP curves were 0.70 (0.64-0.76) and 0.89 (0.80-0.99), respectively.
Conclusion: In patients with COVID-19, fever, leukocytosis, and an elevated CRP were associated with severe outcomes. Leukocytosis and CRP on arrival may predict poor outcomes.
Keywords: COVID-19; CRP; Critically ill; Fever; Luekocytosis; Meta-analysis; Prediction; White blood cell.
Copyright © 2020 Elsevier B.V. All rights reserved.