Sanaka M, Takano K, Shimakura K, Koike Y, Mineshita S
Department of Preventive Medicine, Tokyo Medical and Dental University, Japan.
Nephron. 1996;73(2):137-44. doi: 10.1159/000189030.
The present study was conducted to develop a special formula to predict creatinine clearance (CCr, ml/min) in the elderly with chronic muscle atrophy using serum albumin (Alb, g/dl). We obtained 90 data sets including actual body weight (BW, kg), urinary creatinine excretion (UCr, mg/24 h), serum creatinine (SCr, mg/dl), and Alb from 90 inpatients aged 60-92 years. Regression equations were determined between the dependent variable of UCr/BW and the explanatory variable of Alb as follows: For males Ucr/BW = 2.695 Alb + 4.665 For females Ucr/BW = 1.827 Alb + 4.146. Then, the new predictive formula was derived from the equations: For males CCr = (19Alb + 32)BW/(100 x SCr) For females CCr = (13Alb + 29)BW/(100 x SCr). Evaluations for the predictive error (predicted CCr- measured CCr) showed that the new formula could provide more accurate and less biased estimates of CCr than the Cockcroft and Gault formula could, even in patients with renal insufficiency and in those with Alb < or = 2.8 g/dl.
本研究旨在开发一种特殊公式，用于使用血清白蛋白（Alb，g/dl）预测患有慢性肌肉萎缩的老年人的肌酐清除率（CCr，ml/min）。我们从90名年龄在60 - 92岁的住院患者中获得了90组数据集，包括实际体重（BW，kg）、尿肌酐排泄量（UCr，mg/24 h）、血清肌酐（SCr，mg/dl）和Alb。确定了UCr/BW作为因变量与Alb作为解释变量之间的回归方程如下：男性：Ucr/BW = 2.695Alb + 4.665；女性：Ucr/BW = 1.827Alb + 4.146。然后，从这些方程推导出新的预测公式：男性：CCr = (19Alb + 32)BW/(100 x SCr)；女性：CCr = (13Alb + 29)BW/(100 x SCr)。对预测误差（预测的CCr - 测量的CCr）的评估表明，即使在肾功能不全患者以及Alb≤2.8 g/dl的患者中，新公式也能比Cockcroft和Gault公式提供更准确、偏差更小的CCr估计值。
