Comparison of predictive performances of pneumonia scores, qSOFA and 4C mortality score in patients hospitalized for COVID-19: a retrospective study

Several scores have been validated for predicting mortality in community-acquired pneumonia (CRB-65, CURB-65, PSI), sepsis (qSOFA) and COVID-19 (4C Mortality Score). However, performances of theses scores in pneumonia and COVID-19 as well as the impact of Omicron as predominating virus mutation on prediction performances remained to be determined.In this single-center retrospective study, the prognostic value of different scores in the periods pre-omicron (03/20-12/2021) were compared to those of the omicron period (01/2022-03/2023) in patients hospitalized for COVID-19.Of 622 study patients, 348 (56.2%) had pneumonia. Mortality was 16.2% (101 patients), with a significant difference between periods (71 (23.1%) versus 30 (9.5%), p<0.001). In patients with pneumonia, predictive performance of PSI was significantly better compared to CRB-65, CURB-65 and qSOFA (differences between AUCs PSI vs CRB-65 0.066 (95% CI 0.02-0.11, p=0.006); PSI vs CURB-65 0.047 (95% CI 0-0.09, p=0.042); PSI vs qSOFA 0.124 (95% CI 0.05-0.2, p=0.001) but not when compared to the 4C Mortality Score. There was no difference in predictive performances between periods for any score, also after excluding patients with treatment restrictions.All scores had good predictive performances for low risk and moderate for mortality in patients with COVID-19, with PSI being the best predictor for patients with pneumonia. 4C Mortality Score was a good predictior in all patients. There were no significant differences between the predictive performances of the scores in pre-omicron and omicron periods of the pandemic.