Background: Chest radiograph (CXR) is typically obtained early in patients admitted with COVID-19 and may help guide prognosis and initial management decisions.

Objective: To assess the performance of a severity scoring system on admission CXR in predicting hospital outcomes in patients admitted with COVID-19.

Methods: This retrospective study included 240 patients (142 men, 98 women; median age 65 [50-80]) admitted to the hospital from 3/16/2020 to 4/30/2020 with COVID-19 confirmed by real-time reverse transcriptase-polymerase chain reaction and who underwent CXR within 24 hours of admission. Three chest radiologists and three radiology residents independently scored patients' admission CXR using a 0 to 24-point composite scale (sum of 0-3 scores for extent and severity of disease in upper and lower lung zones). The score's interrater reliability score was assessed using Kendall's W statistic. The six readers' scores were averaged for further analyses. Demographic variables, clinical characteristics, and admission laboratory values were collected from the electronic medical record. ROC analysis was performed to assess the association of CXR severity and mortality. Additional univariable associations, as well as multivariable logistic regression models incorporating patient characteristics and laboratory values, were tested between CXR severity and clinical outcomes.

Results: Interrater reliability of the CXR scores ranged from 0.687 to 0.737 for attendings, 0.653 to 0.762 for residents, and 0.575 to 0.666 for all readers combined. Admission CXR score ≥ 10 achieved 53.0% (35/66) sensitivity and 75.3% (131/174) specificity for predicting hospital mortality. Hospital mortality was 44.9% (35/78) in patients with a high-risk admission CXR (score ≥ 10), versus 19.1% (31/162) in patients with a low-risk CXR (score <10) (p<0.001). Admission CXR severity was an independent predictor of death (OR: 1.17, 95% CI 1.10-1.24, p<0.001). CXR severity was a univariable predictor of intubation (OR: 1.14, 95% CI 1.08-1.21, p<0.001) and continuous renal replacement therapy (CRRT) (OR: 1.15, 95% CI 1.04-1.27, p=0.007), though was not associated with these in multivariable models (p>0.05).

Conclusion: For patients admitted with COVID-19, a simple admission CXR severity score may help predict hospital mortality, intubation, and CRRT.

Clinical Impact: CXR may assist risk assessment and clinical decision making early in the disease course of COVID-19.