动静脉同步换血换血量与胆红素水平关系的回归分析

目的　研究探讨高胆红素血症外周动静脉同步换血换血量和总胆红素(TSB)下降关系的一般适用的回归方程。方法　对2 0例采用外周动静脉同步换血的高胆红素血症患儿,按换血前总胆红素水平的数值高低分两组,对每换血5 0ml/kg进行总胆红素检测,并对方程回归参数的95 %置信区间进行分析和检验。结果　高胆红素血症患儿每换血5 0ml/kg前后总胆红素的平均环比换出率k稳定在2 0～2 3 %之间,差异性对比无显著意义,P >0. 5 ;每增加5 0ml/kg换血的总胆红素累积换出率的差异性比较均有非常显著意义,P <0.0 1 ;经15 0ml/kg换血后总胆红素的累积换出率约为5 0 %。换血量(X)与总胆红素浓度(Y)呈线性相关关系,两组的相关系数分别为rⅠ=0.93 5 6和rⅡ=0 8798,P <0 .0 0 1 ;线性回归方程分别为YⅠ=5 2 3 0 7- 1 86X及YⅡ=3 60 83 -1 2 8X ,回归决定系数分别为RⅠ2 =0. 875 4及RⅡ2 =0 . 774 0 ,回归效果显著,P <0. 0 1 ,回归方程具有统计学意义。结论　回归方程的适用范围由回归参数95 %置信水平区间决定,利用线性插值原理,更能反映换血量与总胆红素浓度水平关系并可一般性用于临床预测和控制指导的回归方程形式为Y =Y0 -aX ,其中:Y0 为换血前总胆红素浓度值;当Y0 ≤4 0 0时,可近似取a =1 3 0 ;当Y0 =4 0 0～5 0 0时,可?

Regression analysis on the relationship between bilirubin volume and exchange transfusion volume

Objective To study the regression equation applicable to the relationship between the decrease of TSB volume and the exchange transfusion volume via peripheral vessels. Methods Divide 20 newly born infants into two groups with hyperbilirubinaemia according to the level of their TSB volume before exchange transfusions are performed. Measure the TSB after exchanged 50 ml/kg, examine and analyze the 95% of the regression coefficient's confidence interval. Results For children with hyperbilirubinaemia, average clearance rate of TSB, i.e.k, before and after exchanged per 50 ml/kg, varies permanently between 20% and 30%. This comparison on diversity is of no great significance with P>0.5. While the diversity comparison of the accumulative clearance rate of TSB after every exchanged per 50 ml/kg is very significant with P<0.01. The accumulative clearance rate of TSB is about 50% after exchanged 150 ml/kg. The exchange transfusion volume (X) is in linear relationship with TSB volume (Y). The coefficients for the two groups are r_Ⅰ=0.9356 and r_Ⅱ=0.8798 respectively with P<0.001. The linear regression equations are Y_Ⅰ=523.07-1.86X and Y_Ⅱ=360.83-1.28X respectively. The decisive regression coefficients are R_Ⅰ~2=0.8754 and R_Ⅱ~2=0.7740 respectively with P<0.01. With marked regression effect, the regression equation is meaningful in terms of statistics. Conclusion The range within which the regression equation is applicable depends on the 95% of the regression coefficient's confidence interval. The relationship between exchange transfusion volume and TSB volume can be better illustrated by employing the theory of linear interpolation. And the regression equation generally applicable to clinic anticipation and directing control takes the form of Y=Y_0-aX, thereof Y_0 equals the TSB volume before the exchange transfusion. When Y_0≤400, a can be considered as 1.30 approximately. When Y_0=400～500, a can be 1.55 approximately. When Y_0≥400, a can be 1.85 approximately. As the exchange transfusion volume increases by 50 ml/kg, the exchange TSB volume decreases at a ratio of (1-k).