Abstract P1037: Comparison of the Updated Sequential Organ Failure Assessment (SOFA-2) Scores and Original SOFA Scores for in-Hospital Mortality Prediction in Cardiac Arrest Clinical Trials
Abstract
Background: The Sequential Organ Failure Assessment (SOFA) score is commonly used in critical care to quantify the severity of organ dysfunction and predict clinical outcomes. Given the evolution of critical care practices, pharmacotherapy, and medical devices, an updated SOFA score (SOFA-2) was developed to more accurately reflect contemporary interventions. We evaluated whether SOFA-2 performs better than SOFA-1 as a predictor of in-hospital mortality using data from two completed cardiac arrest trials: Thiamine as metabolic resuscitator after out-of-hospital cardiac arrest (THACA) and Thiamine as metabolic resuscitator after in-hospital cardiac arrest (THICA). Research Question: Does SOFA-2, compared to SOFA-1 score, improve prediction of in-hospital mortality in cardiac arrest clinical trials? Goal: To compare the performance of SOFA-2 relative to SOFA-1 for discriminating in-hospital mortality and assess differences in organ-specific subscores between SOFA-1 and SOFA-2 in the THICA and THACA cohorts. Methods: Total SOFA scores and organ-specific subscores were calculated for both cohorts. To assess discrimination for in-hospital mortality, area under the receiver operating characteristic curve (AUROC) was compared between SOFA-1 and SOFA-2 at baseline, 24, 48, and 72 hours using tests of equality of ROC areas. Correlation between scoring systems was assessed using Pearson’s correlation coefficient. Patients who died prior to the time point of interest were excluded from analysis at that time point. Results: Median SOFA-2 scores were significantly lower than SOFA-1 in both THACA (9.5 [7, 12] vs. 12 [9, 13], p<0.001) and THICA (11.5 [10, 13.5] vs. 12.5 [10, 14], p=0.002) (Figure 1). Lower SOFA-2 values were consistently seen across most individual organ subscores and time points, except for coagulation and a transiently higher baseline renal subscore in THICA. SOFA-2 showed comparable discrimination to SOFA-1 for in-hospital mortality at all timepoints, with the strongest discrimination at the latest time point present at 72 and 48 hours for THACA and THICA respectively (Figure 2, Table 1). Conclusion: SOFA-2 consistently yielded significantly lower total and individual organ subscores than SOFA-1 but performed similarly with regard to in-hospital mortality.