Prediction of in-hospital recurrence and false-negative results in patients with COVID-19 by red blood cell values on admission.

The clinical characteristics and risk factors of patients with coronavirus disease 2019 (COVID-19) with re-positive or false-negative test results have so far remained to be determined. The present study provides a cross-sectional observational study on 134 hospitalized patients selected from Huoshenshan Hospital (Wuhan, China) using cluster sampling. A total of 68 patients had reduced red blood cell (RBC) counts, 55 a decrease in the hemoglobin concentration (HBC) and 73 a decline in hematocrit (HCT). The false-negative rate of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) RNA detection in pharyngeal swab specimens was 18.7%. The absolute lymphocyte count (ALC), RBC, HBC and HCT levels in false-negative patients were significantly higher than those in patients who tested positive for viral nucleic acids. Multivariate logistic regression analysis indicated that RBC [odds ratio (OR)=0.43, 95% CI: 0.18-0.99], HBC (OR=0.97, 95% CI: 0.94-0.99) and ALC (OR=0.43, 95% CI: 0.20-0.91) were the factors influencing the negative testing results for viral nucleic acid. The rate of re-positive patients was 16.4%. The white blood cell, RBC, HBC and HCT values in re-positive patients were lower than those in non-re-positive patients. The median (interquartile range) values for RBC, HBC and HCT of male re-positive patients were 3.95 (3.37, 4.2) x10/l, 123 (103, 133) g/l and 36.6 (31.1, 39.2)%, respectively, while the RBC, HBC and HCT of female re-positive patients were 3.54 (3.13, 3.74) x10/l, 115 (102, 118) g/l and 34.2 (28.5, 34.9)%, respectively. It was determined that RBC, HBC and HCT values had moderate accuracy in predicting SARS-CoV-2 recurrence in patients with COVID-19 using receiver operating curve analysis. The present study suggested that RBC may have an important role in the pathogenesis of COVID-19.