Search for a command to run...
Objective This observational cohort study aimed to identify risk factors for acute kidney injury (AKI) related to fulminant myocarditis (FM) and for the progression of AKI to chronic kidney disease (CKD), as well as to develop a risk prediction model to help improve the renal prognosis in FM patients. Methods Clinical data were collected for FM patients treated at Central China Fuwai Hospital between December 1, 2018 and June 30, 2025. Patients were categorized into AKI and non-AKI groups, and surviving AKI patients were followed for at least 3 months to observe CKD progression. The logistic regression model was used to analyze the risk factors for FM-associated AKI and its progression to CKD. A receiver operator characteristic (ROC) curve was drawn to evaluate the performance of the clinical risk factor model. Results Of the 408 FM patients included in this study, 201 (49.2%) exhibited FM-associated AKI. Male gender, elevated baseline N-terminal pro b-type natriuretic peptide (NT-pro BNP) and procalcitonin levels and reduced left ventricular ejection fraction were identified as risk factors for FM-associated AKI ( P < 0.05). The clinical risk prediction model including the above factors showed excellent performance for predicting FM-associated AKI (area under curve [AUC] = 0.917, 95% confidence interval [CI]: 0.875–0.960, P < 0.001). Of 188 patients who survived to discharge in the FM-associated AKI group, 66 (35.1%) progressed to CKD. Independent risk factors for progression to CKD included delayed treatment and elevated serum creatinine at discharge. The clinical risk prediction model including these factors exhibited excellent performance for predicting FM-associated AKI to CKD (AUC = 0.898, 95% CI: 0.817–0.979, P < 0.001). Conclusion FM patients have a high risk of renal complications. Patients with FM-associated AKI are also at risk for developing CKD. A logistic prediction model based on independent risk factors can predict poor renal prognosis in FM-associated AKI patients, but this prediction model represents a preliminary exploration and requires external validation for further clinical application.