ABSTRACT
BACKGROUND: The clinical outcomes and differential predictive factors for transient and persistent sepsis-associated AKI (SA-AKI) currently remain unclear. Therefore, using a large international database, this study aimed to compare the in-hospital mortality outcomes and determine the clinical factors for differentiating transient and persistent SA-AKI. MATERIALS AND METHODS: Patients with SA-AKI were identified from the Medical Information Mart for Intensive Care IV (MIMIC IV) database and divided into a persistent SA-AKI group and a transient SA-AKI group. In-hospital mortality rates between the two groups were compared. Logistic regression was performed to explore the risk factors, with the establishment of a clinical predictive model for transient or persistent SA-AKI. RESULTS: Mortality rate of the persistent SA-AKI group was higher than of the transient SA-AKI group (p < 0.001). Persistent SA-AKI was an independent risk factor for in-hospital mortality. Factors including creatinine level at inclusion (OR 1.49, 95% CI 1.43 - 1.56, p < 0.001), changes of creatinine level (mg/dL) within 24 hours (OR 4.25, 95% CI 3.74 - 4.86, p < 0.001), continuous renal replacement therapy (CRRT) performed within 48 hours (OR 4.20, 95% CI 2.87 - 6.17), lactate level (mmol/L) (OR 1.18, 95% CI 1.05 - 1.12, p < 0.001) were independent risk factors for persistent SA-AKI. When incorporating these factors, the predictive model for persistent SA-AKI demonstrated good C-indexes of 0.80 (95% CI: 0.80 - 0.82) and 0.81 (95% CI: 0.79 - 0.83) in the training and validation cohorts, respectively, though the goodness-of-fit was poor. CONCLUSION: Persistent SA-AKI has a higher risk of in-hospital mortality than transient SA-AKI, whereby persistent SA-AKI is also an independent predictor of in-hospital mortality. The nomogram established by predictive factors can be applied in clinical practice to predict persistent SA-AKI.
