Accuracy of conventional disease severity scores in predicting COVID-19 ICU mortality: retrospective single-center study in Turkey

BACKGROUND: Disease severity scores are important tools for predicting mortality in intensive care units (ICUs), but conventional disease severity scores may not be suitable for predicting mortality in coronavirus disease-19 (COVID-19) patients. OBJECTIVE: Compare conventional disease severity scores for discriminative power in ICU mortality. DESIGN: Retrospective cohort SETTING: Intensive care unit in tertiary teaching and research hospital. PATIENTS AND METHODS: COVID-19 patients who were admitted to our ICU between 11 March 2020 and 31 December 2021 were included in the study. Patients who died within the first 24 hours were not included. SAPS II, APACHE II and APACHE 4 scores were calculated within the first 24 hours of ICU admission. A receiver operating characteristics (ROC) analysis was performed for discriminative power of disease severity scores. MAIN OUTCOME MEASURE: ICU mortality SAMPLE SIZE AND CHARACTERISTICS: 510 subjects with median (interquartile percentiles) age of 65 (56-74) years. RESULTS: About half (n=250, 51%) died during ICU stay. Three disease severity scores had similar discriminative power, the area under the curve (AUC), SAPS II (AUC 0.79), APACHE II (AUC 0.76), APACHE 4 (AUC 0.78) (P<.001). Observed mortality was higher than predicted mortality according to conventional disease severity scores. CONCLUSION: Conventional disease severity scores are good indicators of COVID-19 severity. However, they may underestimate mortality in COVID-19. New scoring systems should be developed for mortality prediction in COVID-19. LIMITATION: A single-center study CONFLICT OF INTEREST: None.