IgA肾病联合诊断数学模型的建立与验证

Establishment and Validation of the Mathematical Model of Combined Diagnosis for IgA Nephropathy

目的建立并评价实验室检验指标联合诊断IgA肾病(IgA nephropathy,IgAN)数学模型,探索无创性诊断IgA肾病的方法。方法采集并分析解放军总医院2013年至2015年行肾活检病理学检查的588例住院患者(200例IgAN,388例Non-IgAN)的实验室检查数据。结果 t检验或Mann-Whitney U秩和检验分析表明,在31项实验室检查指标中有23项在IgAN与Non-IgAN中差异具有统计学意义(P<0.05)。多参数Logistic回归分析表明,血清白蛋白(ALB)、血清肌酐(CREA)、血清总胆固醇(CHO)、血清纤维蛋白(FIB)、血清D-二聚体(D-D)、血清IgA(s IgA)、血清IgG(s IgG)7项实验室检查指标对IgAN的诊断具有意义。用这7项指标建立联合诊断数学模型,该模型诊断将单一指标诊断IgAN的ROC曲线下面积(AUC)从0.747从提高到0.890。模型验证结果表明,诊断IgAN与Non-IgAN的总准确度、灵敏度、特异性、阳性预测值、阴性预测值分别为0.813、0.737、0.849、0.700、0.871。结论实验室检查项目ALB、CREA、CHO、FIB、D-D、IgA和IgG具有联合诊断IgAN的能力,基于多项实验室检查、联合诊断和数据分析的方法可用于IgAN的诊断,具有较好的临床应用价值。

Objective Establishing and validating the mathematical model of combined laboratory examinations for classifying IgA nephropathy（IgAN） from Non-IgA nephropathy（Non-IgAN）,and studying the method of non-invasive diagnosis for IgAN. Methods The laboratory data of 588 inpatients（200 IgAN,388 NonIgAN） who received renal biopsy for pathological examinations in PLA general hospital from 2013 to 2015 were collected and analyzed. Results The results of T-test or Mann-Whitney U analysis suggested that 23 out of 31 laboratory tests had significant difference（P〈0. 05） between IgAN and Non-IgAN. Multivariate Logistic regression analysis indicated 7 tests of ALB, CREA, CHO, FIB, D-D, IgA and IgG had significance for diagnosing IgAN. The mathematical model with combination of these 7 indicators could elevate the area under ROC curve（AUC） from 0. 747 to 0. 890. Validation results show ed that the accuracy,sensitivity,specificity,positive predictive value and negative predictive value of the model were 0. 813,0. 737,0. 849,0. 700,0. 871,respectively,for diagnosing IgAN from Non-IgAN. Conclusion Laboratory tests of ALB,CREA,CHO,FIB,D-D,IgA and IgG had diagnostic pow er of discerning IgAN. Our approach of using multiple laboratory examinations,combining with data analysis could had prospective clinical application value.