Objective To investigate the risk factors for idiopathic membranous nephropathy (IMN) complicated by acute kidney injury (AKI) based on propensity score matching (PSM).
Methods A total of 310 IMN patients admitted between January 2020 and March 2025 were retrospectively enrolled and divided into an AKI group (n = 80) and a non-AKI group (n = 230). Patients in the non-AKI group were matched with those in the AKI group at a 1:1 ratio using PSM, resulting in 80 matched pairs. The general data of the patients were collected. The primary outcome measures included serum creatinine, estimated glomerular filtration rate (eGFR), uric acid, P-selectin, and urinary kidney injury molecule-1 (KIM-1) levels. Logistic regression analysis was performed to identify factors associated with risks for IMN complicated by AKI, and a nomogram prediction model was subsequently established.
Results After PSM, all 5 primary outcome measures differed significantly between the AKI and non-AKI groups (all P < 0.05). According to multivariate logistic regression analysis, 4 outcome measures, including serum creatinine, eGFR, serum uric acid, and serum P-selectin, were all identified as factors influencing the occurrence of AKI in patients with IMN (all P < 0.05). A combined prediction model (a nomogram model) incorporating these 4 variables yielded an area under the curve (AUC) of the receiver operating characteristic curve (ROC) of 0.945 (95% CI: 0.909-0.981). The calibration curve demonstrated a good fit between the predicted and the ideal curves. Findings from the Hosmer-Lemeshow goodness-of-fit test supported adequate calibration of the model (χ² = 7.366, P = 0.498). Shapley additive explanations (SHAP) importance ranking analysis showed that serum P-selectin was an important influencing factor for AKI in patients with IMN, and the importance of serum uric acid, eGFR, and serum creatinine was secondary.
Conclusion Serum creatinine, eGFR, blood uric acid, and serum P-selectin are independent influencing factors for patients with acute kidney injury, and they have certain predictive value.