Predictors of Intensive Care Unit admission in patients with coronavirus disease 2019 (COVID-19)

Italy is currently experiencing an epidemic of coronavirus disease 2019 (Covid-19). Aim of our study is to identify the best predictors of Intensive Care Unit (ICU) admission in patients with Covid-19. We examined 28 patients admitted to the Emergency Department (ED) and subsequently confirmed as cases of Covid-19. Patients received, at the admission to the ED, a diagnostic work-up including: patient history, clinical examination, an arterial blood gas analysis (whenever possible performed on room air), laboratory blood tests, including serum concentrations of interleukin 6 (IL-6), lung ultrasound examination and a computed tomography (CT) scan of the thorax. For each patient, as gas exchange index through the alveolocapillary membrane, we determined the alveolar-arterial oxygen gradient (AaDO(2)) and the alveolar-arterial oxygen gradient augmentation (AaDO(2) augmentation). For each patient, as measurement of hypoxemia, we determined oxygen saturation (SpO(2)), partial pressure of oxygen in arterial blood (PaO2), PaO2 deficit and the ratio between arterial partial pressure of oxygen by blood gas analysis and fraction of inspired oxygen (P/F). Patients were assigned to ICU Group or to Non-ICU Group basing on the decision to intubate. Areas under the curve (AUC) and receiver operating characteristic (ROC) curve were used to compare the performance of each test in relation to prediction of ICU admission. Comparing patients of ICU Group (10 patients) with patients of Non-ICU Group (18 patients), we found that the first were older, they had more frequently a medical history of malignancy and they were more frequently admitted to ED for dyspnea. Patients of ICU Group had lower oxygen saturation, PaO2, P/F and higher heart rate, respiratory rate, AaDO2, AaDO2 augmentation and lactate than patients of Non-ICU Group. ROC curves demonstrate that age, heart rate, respiratory rate, dyspnea, lactate, AaDO(2), AaDO(2) augmentation, white blood cell count, neutrophil count and percentage, fibrinogen. C-reactive protein, lactate dehydrogenase, glucose level, international normalized ratio (INR), blood urea and IL-6 are useful predictors of ICU admission. We identified several predictors of ICU admission in patients with Covid-19. They can act as fast tools for the early identification and timely treatment of critical cases since their arrival in the ED.