ObjectiveTo evaluate the diagnostic value of various severity assessment scoring systems for sepsis after cardiac surgery and the predictive value for long-term prognosis.MethodsThe clinical data of patients who underwent cardiac sugeries including coronary artery bypass grafting (CABG) and (or) valve reconstruction/valve replacement were extracted from Medical Information Mark for Intensive Care-Ⅲ (MIMIC-Ⅲ). A total of 6 638 patients were enrolled in this study, including 4 558 males and 2 080 females, with an average age of 67.0±12.2 years. Discriminatory power was determined by comparing the area under the receiver operating characteristic (ROC) curve (AUC) for each scoring system individually using the method of DeLong. An X-tile analysis was used to determine the optimal cut-off point for each scoring system, and the patients were grouped by the cut-off point, and Kaplan-Meier curves and log-rank test were applied to analyze their long-term survival.ResultsCompared with the sequential organ failure assessment (SOFA) score, acute physiology score-Ⅲ (APS-Ⅲ, P<0.001), the simplified acute physiology score-Ⅱ (SAPS-Ⅱ, P<0.001) and logistic organ dysfunction score (LODS, P<0.001) were more accurate in distinguishing sepsis. Compared with the non-septic group, the 10-year overall survival rate of the septic group was lower (P<0.001). Except for the systemic inflammation response score (SIRS) system, the 10-year overall survival rates of patients in the high risk layers of SOFA (HR=2.50, 95%CI 2.23-2.80, P<0.001), SAPS (HR=2.93, 95%CI 2.64-3.26, P<0.001), SAPS-Ⅱ (HR=2.77, 95%CI 2.51-3.04, P<0.001), APS-Ⅲ (HR=2.90, 95%CI 2.63-3.20, P<0.001), LODS (HR=2.17, 95%CI 1.97-2.38, P<0.001), modified logistic organ dysfunction score (MLODS, HR=2.04, 95%CI 1.86-2.25, P<0.001) and the Oxford acute severity of illness score (OASIS, HR=2.37, 95%CI 2.16-2.60, P<0.001) systems were lower than those in the low risk layers.ConclusionCompared with SOFA score, APS-Ⅲ score may have higher value in the diagnosis of sepsis in patients who undergo isolated CABG, a valve procedure or a combination of both. Except for SIRS scoring system, SOFA, APS-Ⅲ, SAPS, SAPS-Ⅱ, LODS, MLODS and OASIS scoring systems can be applied to predict the long-term outcome of patients after cardiac surgery.