CAJ | 학술논문

目的：评估用尿蛋白肌酐比( protein/creatinine ratio,P/C ratio)预测24 h尿蛋白定量的有效性。方法：选取在我院肾内科就诊的慢性肾脏病( chronic kidney disease,CKD)患者138例,共199份标本,根据肾小球滤过率、血清白蛋白水平分为若干亚组。将患者尿蛋白肌酐比与24 h尿蛋白定量进行相关性分析,采用ROC曲线分析尿蛋白肌酐比相对于24 h尿蛋白定量为0.5 g、1.0 g及3.5 g的最佳诊断界点。同时采用线性回归模型建立24 h尿蛋白定量预测方程。结果：尿蛋白肌酐比与24 h尿蛋白定量有良好相关性(r=0.812)。其相关系数在CKD1~4期分别是0.791,0.925,0.872,0.893,而在CKD 5期其相关系数则较低(0.450)。高血清白蛋白亚组(r=0.817)比低血清白蛋白亚组(r=0.652)显示出更高的相关性。应用ROC曲线分析,尿蛋白肌酐比相对应24 h尿蛋白定量0.5 g/24 h、1.0 g/24 h及3.5 g/24 h诊断界点分别为0.45 mg/mg (敏感性0.95；特异性0.88),0.81 mg/mg(敏感性0.97；特异性0.77)和3.92 mg/mg(敏感性0.97；特异性0.77)时敏感性及特异性最优。用线性回归方法建立回归方程：24 h尿蛋白定量( g)=5.543+0.422×尿蛋白肌酐比( mg/mg)-0.104×血清白蛋白( g/L)。结论：尿蛋白肌酐比与24 h尿蛋白定量有较好相关性,可以替代24 h尿蛋白定量,但在CKD5期和低血清白蛋白患者其准确性下降。
목적：평고용뇨단백기항비( protein/creatinine ratio,P/C ratio)예측24 h뇨단백정량적유효성。방법：선취재아원신내과취진적만성신장병( chronic kidney disease,CKD)환자138례,공199빈표본,근거신소구려과솔、혈청백단백수평분위약간아조。장환자뇨단백기항비여24 h뇨단백정량진행상관성분석,채용ROC곡선분석뇨단백기항비상대우24 h뇨단백정량위0.5 g、1.0 g급3.5 g적최가진단계점。동시채용선성회귀모형건립24 h뇨단백정량예측방정。결과：뇨단백기항비여24 h뇨단백정량유량호상관성(r=0.812)。기상관계수재CKD1~4기분별시0.791,0.925,0.872,0.893,이재CKD 5기기상관계수칙교저(0.450)。고혈청백단백아조(r=0.817)비저혈청백단백아조(r=0.652)현시출경고적상관성。응용ROC곡선분석,뇨단백기항비상대응24 h뇨단백정량0.5 g/24 h、1.0 g/24 h급3.5 g/24 h진단계점분별위0.45 mg/mg (민감성0.95；특이성0.88),0.81 mg/mg(민감성0.97；특이성0.77)화3.92 mg/mg(민감성0.97；특이성0.77)시민감성급특이성최우。용선성회귀방법건립회귀방정：24 h뇨단백정량( g)=5.543+0.422×뇨단백기항비( mg/mg)-0.104×혈청백단백( g/L)。결론：뇨단백기항비여24 h뇨단백정량유교호상관성,가이체대24 h뇨단백정량,단재CKD5기화저혈청백단백환자기준학성하강。
Objective:The purpose of our study is to assess the effectiveness of protein/creatinine ( P/C) ratio to predict 24 hour urinary protein. Methods:A total of 199 chronic kidney disease ( CKD) patients were enrolled. All patients were assigned in-to five CKD categories according to the NKFK/DOQI clinical practice guidelines and two subgroups by serum albumin level≥30 g/L and <30 g/L. Correlation coefficients of the urine P/C ratio and 24 hour urinary protein were calculated. Receiver operating charac-teristic curves were used to identify the cutoff values for urine P/C ratio for predicting 24 hour urinary protein“threshold” excretion of 0. 5, 1. 0 and 3. 5 g/24 h. Linear regression model was used to develop an equation to evaluate 24 hour urinary protein. Results:Cor-relation coefficient of the urine P/C ratio and 24 hour urinary protein was 0. 812 (P<0. 001). Correlation coefficients in CKD 1 to 5 categories were 0. 791, 0. 925, 0. 872, 0. 893 and 0. 450, respectively. The higher serum albumin subgroup had a better correlation coefficient (0. 817) than the lower one (0. 652). The urine P/C value of 0. 45 mg/mg (sensitivity 0. 95; specificity 0. 88), 0. 81 mg/mg (0. 97;0. 77) and 3. 92 mg/mg (0. 81;0. 87) were the predictors for 24 hour urinary protein equivalent‘thresholds’ at 0. 5, 1. 0 and 3. 5 g/day. Finally, an equation using the linear regression model was developed:[24 hour urinary protein ( g) =5. 543+0. 422 × urine P/C radio (mg/mg) -0. 104 × serum albumin (g/L)]. Conclusion:Urine P/C ratio can act as a predictor in the evaluation of 24 hour urinary protein. However, the performance varies in high CKD category and low serum albumin subgroups.