目的 研究儿童过敏性紫癜肾炎（Henoch-Schönlein purpura nephritis, HSPN）肾小球损伤进展的影响因素，并构建预测模型。 方法 选取2022年1月—2024年12月在河南中医药大学第一附属医院住院的HSPN患儿259例，根据肾脏穿刺病理报告，将患儿分为低分级组（肾小球损伤Ⅰ~Ⅱ级，61例）和高分级组（肾小球损伤Ⅲ~Ⅵ级，198例），比较两组间的临床指标和病理特征，采用多因素logistic回归构建肾小球损伤进展为高分级的预测模型，并通过受试者操作特征曲线分析、Hosmer-Lemeshow拟合优度检验评价模型的区分度与拟合度。 结果 与低分级组相比，高分级组的尿红细胞计数、尿管型计数、尿小圆上皮细胞计数、尿蛋白/肌酐比值、尿IgG/肌酐比值、尿N-乙酰-β-D-葡萄糖苷酶、纤维蛋白原、白细胞计数和中性粒细胞/淋巴细胞比值均显著升高，白蛋白、白蛋白/球蛋白比值和凝血酶原时间均显著下降，血脂异常、弥漫性系膜增生、肾小球节段性硬化或粘连、肾小管萎缩/间质纤维化及新月体形成的比例均显著升高（均P<0.05）。多因素logistic回归分析显示，小圆上皮细胞计数、尿蛋白/肌酐比值、纤维蛋白原水平、白蛋白水平、中性粒细胞/淋巴细胞比值、血脂异常是肾小球从低分级进展为高分级的影响因素（均P<0.05），并基于这些因素建立了预测模型。Hosmer-Lemeshow检验显示该模型拟合优良（P=0.977）。受试者操作特征曲线分析显示，该模型的曲线下面积为0.818（95%CI：0.766~0.863），灵敏度为71.2%，特异度为80.3%（P<0.05）。 结论 HSPN肾小球分级较重的患儿临床和病理特征更为严重，基于小圆上皮细胞计数、尿蛋白/肌酐比值、纤维蛋白原水平、白蛋白水平、中性粒细胞/淋巴细胞比值、血脂异常等指标构建的预测模型可以较好地预测肾小球损伤进展为高分级（Ⅲ~Ⅵ级），可在临床中应用。

Objective To investigate the influencing factors for the progression of glomerular injury in children with Henoch-Schönlein purpura nephritis (HSPN) and to develop a predictive model. Methods A total of 259 children with HSPN admitted to the First Affiliated Hospital of Henan University of Traditional Chinese Medicine from January 2022 to December 2024 were retrospectively enrolled. Based on renal biopsy pathology reports, the children were classified into a low-grade group (grade Ⅰ-Ⅱ glomerular injury; n=61), and a high-grade group (grade Ⅲ-Ⅵ glomerular injury; n=198). Clinical indicators and pathological characteristics were compared between the two groups. A multivariable logistic regression model was constructed to predict progression to high-grade injury. The model's discrimination and calibration were evaluated using receiver operating characteristic curve analysis and the Hosmer-Lemeshow goodness-of-fit test. Results Compared with the low-grade group, the high-grade group showed significantly higher urinary red blood cell count, urinary cast count, urinary small round epithelial cell count, urine protein-to-creatinine ratio, urinary IgG-to-creatinine ratio, urinary N-acetyl-β-D-glucosaminidase, fibrinogen, white blood cell count, and neutrophil-to-lymphocyte ratio, while albumin, albumin-to-globulin ratio, and prothrombin time were significantly decreased (all P<0.05). The proportions of patients with dyslipidemia, diffuse mesangial hyperplasia, glomerular segmental sclerosis or adhesion, tubular atrophy/interstitial fibrosis, and crescent formation were also significantly higher (all P<0.05). Multivariable logistic regression identified urinary small round epithelial cell count, urine protein-to-creatinine ratio, fibrinogen level, albumin level, neutrophil-to-lymphocyte ratio, and dyslipidemia as factors associated with progression from low to high-grade injury (all P<0.05). A prediction model was constructed based on these variables. The Hosmer-Lemeshow test indicated excellent calibration (P=0.977). Receiver operating characteristic analysis showed an area under the curve of 0.818 (95%CI: 0.766-0.863), with a sensitivity of 71.2% and a specificity of 80.3% (P<0.05). Conclusions Children with higher-grade glomerular injury from HSPN exhibit more severe clinical and pathological manifestations. The prediction model incorporating urinary small round epithelial cell count, urine protein-to-creatinine ratio, fibrinogen level, albumin level, neutrophil-to-lymphocyte ratio, and dyslipidemia demonstrates good predictive performance for identifying progression to high-grade (Ⅲ-Ⅵ) glomerular injury and has potential for clinical application.