Predictive Performance of Rapid Scoring Scales in Intensive Care Patients with COVID-19

COVID-19 remains one of the most significant causes of hospital admissions and deaths, frequently associated with septic complications. Reliable and simple prognostic tools are required to identify patients at high risk of developing sepsis and death early on. Aim of the study. To evaluate the prognostic ability of the REMS, NEWS, NEWS2, qSOFA, and SIRS criteria in predicting sepsis, septic shock, and mortality in patients with COVID-19. Materials and methods. A retrospective study was performed on 870 patients with confirmed COVID-19 who were hospitalized in anesthesiology and intensive care units. REMS, NEWS, NEWS2, qSOFA, and SIRS scores were calculated for each patient upon admission. The predictive ability of the scales was evaluated using ROC analysis, temporal associations were assessed using the Cox proportional hazards model, and the cumulative risk of outcome was assessed using the Kaplan–Meier method. Results. The REMS scale demonstrated satisfactory predictive ability for mortality and septic shock (AUC = 0.780 and 0.724, respectively) and unsatisfactory for sepsis (AUC = 0.677), p  < 0.0001. The NEWS, NEWS2, qSOFA, and SIRS scales showed unsatisfactory results for all outcomes. The Kaplan–Meier analysis showed that patients with high REMS scores developed sepsis or septic shock earlier and had shorter survival ( p  < 0.0001). The Cox model identified REMS as the scale with the highest risk ratio (RR 1.215; 95% CI 1.178–1.254) and concordance index (C = 0.656). Conclusion. The REMS, NEWS, NEWS2, qSOFA, and SIRS rapid scales have a statistically significant prognostic value, but only the REMS provides a satisfactory accuracy in predicting septic shock and death in patients with COVID-19. The study was limited by its single-center and retrospective design.