Mortality Predictive Value of APACHE II and SOFA Scores in COVID-19 Patients in the Intensive Care Unit

Abstract

Background. COVID-19 pandemic has become a global dilemma since December 2019. Are the standard scores, such as acute physiology and chronic health evaluation (APACHE II) and sequential organ failure assessment (SOFA) score, accurate for predicting the mortality rate of COVID-19 or the need for new scores? We aimed to evaluate the mortality predictive value of APACHE II and SOFA scores in critically ill COVID-19 patients. Methods. In a cohort study, we enrolled 204 confirmed COVID-19 patients admitted to the intensive care units at the Imam Khomeini hospital complex. APACHE II on the first day and daily SOFA scoring were performed. The primary outcome was the mortality rate in the nonsurvived and survived groups, and the secondary outcome was organ dysfunction. Two groups of survived and nonsurvived patients were compared by the chi-square test for categorical variables and an independent sample t-test for continuous variables. We used logistic regression models to estimate the mortality risk of high APACHE II and SOFA scores. Result. Among 204 severe COVID-19 patients, 114 patients (55.9%) expired and 169 patients (82.8%) had at least one comorbidity that 103 (60.9%) of them did not survive ( $P=0.002$ ). Invasive mechanical ventilation and its duration were significantly different between survived and nonsurvived groups ( $P\le 0.001$ and $P=0.002$ , respectively). Mean APACHE II and mean SOFA scores were significantly higher in the nonsurvived than in the survived group (14.4 ± 5.7 vs. 9.5 ± 5.1, $P\le 0.001$ , 7.3 ± 3.1 vs. 3.1 ± 1.1, $P\le 0.001$ , respectively). The area under the curve was 89.5% for SOFA and 73% for the APACHE II score. Respiratory diseases and malignancy were risk factors for the mortality rate ( $P=0.004$ and $P=0.007$ , respectively) against diabetes and hypertension. Conclusion. The daily SOFA was a better mortality predictor than the APACHE II in critically ill COVID-19 patients. But they could not predict death with high accuracy. We need new scoring with consideration of the prognostic factors and daily evaluation of changes in clinical conditions.