应用多元回归分析建立纠正高胆红素干扰酶法测定血清总胆固醇结果的校正公式及验证

目的 评价不同浓度胆红素对酶法测定总胆固醇(total cholesterol, TC)的影响，应用多元回归分析建立校正公式并验证其校正效果。方法 收集100 例来自于南京医科大学第一附属医院2020 年12 月门诊或住院患者的新鲜无黄疸血清标本。配制一系列浓度梯度的胆红素溶液，分别加入其中50 例患者的血清标本中，检测各标本总胆红素(total bilirubin，TB)和TC 浓度，分析胆红素水平对TC 检测结果的影响及二者的相关性。应用多元回归分析建立校正公式，并利用另外50 例患者标本验证该校正公式的性能。结果 黄疸实验组与生理盐水对照组间差异具有统计学意义(F=3.947，P<0.01)，标本TC 浓度与TB 浓度呈负相关(r2=0.989 4)；在胆红素加入浓度分别为100，200，300，400，500 和600 μmol/L 的实验组中，TC 浓度的平均偏倚依次为:-6.65%±4.89%，-14.21%±8.89%，-29.00%±13.43%，-31.07%±14.62%，-38.21%±13.95% 和-43.51%±12.69%，且初始浓度TC ＜ 3 mmol/L 的标本产生的平均偏倚均明显高于TC ≥ 3 mmol/L 的标本，差异具有统计学意义(t=9.192 ～ 14.836，均P ＜ 0.01)；设无黄疸血清标本的TC 浓度为Y，其对应人工黄疸标本的TC 实测浓度为Z，TB 浓度为X，经多元回归分析，校正公式为Y =0.002 353·X + 0.973·Z+ 0.000 337·X·Z － 0.013 34；经公式校正后92.67% 的黄疸标本TC 浓度偏倚小于±10%。结论 黄疸患者血清标本中高胆红素水平可导致其TC 的检测结果偏低，偏倚大小与TB 浓度以及初始TC 浓度相关，运用校正公式可有效校正高胆红素对TC 测定的干扰，符合临床要求。

Establish A Regression Formula for Correction of the Interference of High Bilirubin on Total Cholesterol Determined by Enzymatic Method and Its Verification

Objective To investigate the effects of bilirubin on total cholesterol (TC) determined by enzymatic method and establish a regression formula and verify its correction effect. Methods Fresh non-icteric serum samples of 100 outpatients or inpatients in the First Affiliated Hospital of Nanjing Medical University were collected in December 2020. A series of bilirubin solutions were prepared and added to the serum samples of 50 patients, respectively. Total bilirubin (TB) and TC concentration in non-icteric serum and different degree icteric serum samples were measured. The impact of bilirubin on TC and the relationship between bilirubin and TC were analyzed. Multiple regression analysis was used to establish a correction formula, and its correction effect was verified with serum samples of another 50 patients. Results The difference between the icteric experimental groups and the saline control group was statistically significant (F=3.947, P ＜ 0.01), and the TC concentration was negatively correlated with the TB concentration(r2=0.989 4). In the icteric experimental groups with added bilirubin concentrations of 100, 200, 300, 400, 500 and 600 μmol/L, the average biases of TC concentrations were -6.65%±4.89%, -14.21%±8.89%, -29.00%±13.43%, -31.07%±14.62%, -38.21%±13.95% and -43.51%±12.69%, respectively. The average bias of samples with initial TC concentration ＜ 3 mmol/L was significantly higher than that of samples with initial TC concentration ≥ 3 mmol/L, and the differences were statistically significant (t=9.192 ～ 14.836, all P ＜ 0.01). Y and Z represented TC concentration of non-icteric serum and different degree icteric serum，while X represented TB concentration． Formulas for TC measurement were presented:Y = 0.002 353·X + 0.973·Z+ 0.000 337·X·Z-0.013 34. The bias of corrected TC in 92.67% of icteric samples was less than 10%. Conclusion The results indicate that the level of TC measured by enzymatic method was negatively affected by bilirubin. The bias was related to the TB concentration and the initial TC concentration. The formula of mathematical correction of TC results in icteric samples should be suitable for the requirements of clinical laboratory.