Performance features and mortality prediction of the 4C Score early in COVID-19 infection: a retrospective study in Saudi Arabia.

The ISARIC4C consortium developed and internally validated the 4C Score for prediction of mortality only in hospitalized patients. We aimed to assess the validity of the 4C Score in mortality prediction of patients with COVID-19 who had been home isolated or hospitalized.This retrospective cross-sectional study was performed after the first wave of COVID-19. Data of all PCR-positive COVID-19 patients who had been discharged, hospitalized, or died were retrospectively analyzed. Patients were classified into four risk groups according to the 4C Mortality Score. A total of (506) patients were classified as follows: low (57.1%), intermediate (27.9%), high (13%), and very high (2%) risk groups. Clinical, radiological, and laboratory data were significantly more severe in the high and very high-risk groups compared with other groups (p<0.001 for all). Mortality rate was correctly estimated by the model with 71% sensitivity, 88.6% specificity, and area under the curve of 0.9. The mortality rate was underestimated among the very high-risk group (66.2% vs 90%). The odds of mortality were significantly greater in the presence of hypoxia (OR 2.6, 95% CI 1.5 to 4.6, p<0.001) and high respiratory rate (OR 5.3, 95% CI 1.6 to 17.9, p<0.007), C reactive protein (CRP) (OR 3.5, 95% CI 1.8 to 6.8, p<0.001), and blood urea nitrogen (BUN) (OR 1.9, 95% CI 1.3 to 3.1, p<0.002). Other components of the model had non-significant predictions. In conclusion, the 4C Mortality Score has good sensitivity and specificity in early risk stratification and mortality prediction of patient with COVID-19. Within the model, only hypoxia, tachypnea, high BUN, and CRP were the independent mortality predictors with the possibility of overlooking other important predictors.