부산대학교 도서관

Abstract Background Being highly infectious disease, COVID-19 exhausts most of efficient healthcare systems worldwide. Simple and rapid risk stratification methods are mandatory to recognize severe patients. This study aims to highlight the simple available laboratory biomarkers of good predictive value for COVID-19 severity. Results Three hundred fifty-one COVID-19 positive patients admitted to two University Hospitals between the 1st of June and the 31st of July 2020 were retrospectively collected and classified to severe and non-severe COVID-19 patients according to need for ICU admission. All basic laboratory biomarkers at time of admission were recorded. Of included patients, 145 (41.3%) needed ICU admission. Anemia, leukocytosis, lymphopenia, NLR, and PLR together with liver enzymes, INR, ferritin, CRP, and D-dimer were significantly higher in patients needed ICU admission (p < 0.001). However, by applying multivariate logistic regression, only anemia, high NLR, high PLR, and high D-dimer levels showed significant risk for ICU admission with OR equal 3.6 (95% CI 1.8–7.0), 9.0 (95% CI 3.6–22.6), 3.0 (95% CI 1.3–7.1), and 2.5 (95% CI 1.3–4.7), respectively. Conclusion Anemia, increased neutrophil-to-lymphocyte ratio (> 8), platelet-to-lymphocyte ratio (> 192), and D-dimer level (> 0.9 mg\L) at time of admission could be simple available predictors for severe COVID-19 infection requiring ICU admission.