Abstract
BACKGROUND: Serum creatinine, an important diagnostic indicator for acute kidney injury (AKI), was considered to be a risk factor for cardiovascular disease. This study aimed to investigate the significance of postoperative serum creatinine in predicting the prognosis of cardiac surgery patients. METHODS: The Medical Information Mart for Intensive Care III (MIMIC-III) database was used to extract the clinical data. Adult (≥18 years) cardiac surgery patients in the database were enrolled. The correlation of postoperative serum creatinine with lengths of intensive care unit (ICU) stay was analyzed with Spearman correlation, and the association of postoperative serum creatinine with hospital mortality was analyzed with chi-square tests. Multivariable logistic regression was used to identify postoperative serum creatinine as an independent prognostic factor for hospital mortality. RESULTS: A total of 6,001 patients were enrolled in our study, among whom, 108 patients (1.8%) died in the hospital. Non-survivors had much higher postoperative serum creatinine levels (initial: 0.8 vs. 1.2 mg/dl, P < 0.001; maximum: 1.1 vs. 2.8 mg/dl, P < 0.001; minimum: 0.8 vs.1.1 mg/dl, P < 0.001). Positive correlations were observed between postoperative serum creatinine (P < 0.001) and lengths of ICU stay. For all models, postoperative initial creatinine, postoperative maximum creatinine, and postoperative minimum creatinine were all positively associated with hospital mortality (all P < 0.001). The predictive performance of postoperative serum creatinine was moderately good (area under the curve (AUC) for initial creatinine = 0.7583; AUC for maximum creatinine = 0.8413; AUC for minimum creatinine = 0.7063). CONCLUSIONS: This study demonstrated the potential to use postcardiac surgery serum creatinine as an outcome indicator.