Division of Nephrology, Department of Medicine, University Medical Center Groningen, University Hospital Groningen, Groningen, The Netherlands.
Clin J Am Soc Nephrol. 2011 Jul;6(7):1731-8. doi: 10.2215/CJN.08590910.
There is a need for prediction scores that identify individuals at increased risk for developing progressive chronic kidney disease (CKD). Therefore, this study was performed to develop and validate a "renal risk score" for the general population. Design, setting, participants, & measurements For this study we used data from the PREVEND (Prevention of Renal and Vascular ENdstage Disease) study, a prospective population-based cohort study with a median follow-up of 6.4 years. Participants with two or three consecutive estimated GFR (eGFR) measurements during follow-up were included. Participants within the group who had the most renal function decline (top 20% of the total population) and had an eGFR value <60 ml/min per 1.73 m² during follow-up were defined as having progressive CKD. Possible predictors for progressive CKD were selected on the basis of univariable logistic regression analyses.
A final prediction model was built using backward logistic regression analysis. Besides baseline eGFR, the model contained age, urinary albumin excretion, systolic BP, C-reactive protein, and known hypertension. The area under the receiver operating characteristic (ROC) curve was 0.84. We performed internal validation by using a bootstrapping procedure. As expected, after the regression coefficients were corrected for optimism, the area under the ROC curve was still 0.84. For clinical use we divided all predictors in meaningful clinical categories to develop a score chart. The area under the ROC curve was 0.83, indicating the high discriminative value of this model.
Given the high internal validity of this renal risk score, this score can be helpful to identify individuals at increased risk for progressive CKD.
需要有一种预测评分系统，用于识别发生进展性慢性肾脏病（CKD）风险增加的个体。因此，本研究旨在开发和验证一种针对普通人群的“肾脏风险评分”。
设计、地点、参与者和测量：本研究使用了 PREVEND（预防肾脏和血管终末期疾病）研究的数据，这是一项前瞻性的基于人群的队列研究，中位随访时间为 6.4 年。纳入了在随访期间有两次或三次连续估算肾小球滤过率（eGFR）测量值的参与者。在该组中，肾功能下降最多（总人群的前 20%）且在随访期间 eGFR 值<60 ml/min/1.73 m²的参与者被定义为发生进展性 CKD。基于单变量逻辑回归分析选择了可能预测进展性 CKD 的预测因素。
使用向后逻辑回归分析构建了最终的预测模型。除了基线 eGFR，该模型还包含年龄、尿白蛋白排泄、收缩压、C 反应蛋白和已知的高血压。接收者操作特征（ROC）曲线下面积为 0.84。我们通过使用自举程序进行了内部验证。正如预期的那样，在对回归系数进行校正以消除乐观偏差后，ROC 曲线下面积仍为 0.84。为了临床应用，我们将所有预测因素划分为有意义的临床类别，以开发评分图表。ROC 曲线下面积为 0.83，表明该模型具有较高的区分能力。
鉴于该肾脏风险评分具有较高的内部有效性，该评分可有助于识别进展性 CKD 风险增加的个体。
