Prediction of Acute Lung Injury Assessed by Chest Computed Tomography, Oxygen Saturation/fraction of Inspired Oxygen Ratio, and Serum Lactate Dehydrogenase in Patients with COVID-19.

Introduction: In treating acute hypoxemic respiratory failure (AHRF) caused by coronavirus disease 2019 (COVID19), clinicians choose respiratory therapies such as low-flow nasal cannula oxygenation, high-flow nasal cannula oxygenation, or mechanical ventilation after assessment of the patient 's condition. Chest computed tomography (CT) imaging contributes significantly to diagnosing COVID-19 pneumonia. However, the costs and potential harm to patients from radiation exposure need to be considered. This study was performed to predict the quantitative extent of COVID-19 acute lung injury using clinical indicators such as an oxygenation index and blood test results. Methods: We analyzed data from 192 patients with COVID-19 AHRF. Multiple logistic regression was used to determine correlations between the lung infiltration volume (LIV) and other pathophysiological or biochemical laboratory parameters. Results: Among 13 clinical parameters, we identified the oxygen saturation/fraction of inspired oxygen ratio (SF ratio) and serum lactate dehydrogenase (LD) concentration as factors associated with the LIV. In the binary classification of an LIV of >= 20 % or not and with the borderline LD = 2.2 x [SF ratio] -182.4, the accuracy, precision, diagnostic odds ratio, and area under the summary receiver operating characteristic curve were 0.828, 0.818, 23.400, and 0.870, respectively. Conclusions: These data suggest that acute lung injury due to COVID-19 pneumonia can be estimated using the SF ratio and LD concentration without a CT scan. These findings may provide significant clinical benefit by allowing clinicians to predict acute lung injury levels using simple, minimally invasive assessment of oxygenation capacity and biochemical blood tests.