摘要
目的比较由匀相测定法与Friedewald公式法计算得到的低密度脂蛋白胆固醇（LDL-C）值的差异,评估两者在临床应用中的相关性及可能出现的偏差。方法选择1 180例门诊及住院患者,测定空腹血清LDL-C、高密度脂蛋白胆固醇（HDL-C）、总胆固醇（TC）、三酰甘油（TG）水平,以匀相测定法为对比方法,计算法为实验方法,计算相关系数（r）和直线回归方程,对两者之间的预期偏差进行评估;并以实测值为参考值,计算在不同浓度范围LDL-C计算值与实测值的符合率、计算值在给定医学决定水平处的阳性预测值和阴性预测值。结果2种方法的r〉0.97（P〈0.001）,在不同医学决定水平预期偏差均在其可信区间上限与下限之间;以实测值为参考值,多数情况下计算值与实测值的符合率〉85%,仅在3.12-3.64 mmol/L之间,两者符合率较低;在不同医学决定水平,计算值具有较高的阴性预测值（〉98%）和阳性预测值（85%）,仅在3.64 mmol/L处,阳性预测值较低（77.8%）。结论计算法与测定法有良好的相关性,计算法的预期偏差在可接受范围内。在不同的医学决定水平,计算法能较好地区分“阴性”和“阳性”样本,对于实测值在3.12-3.64 mmol/L之间的样本,计算法可能较高地估计了LDL-C的浓度。
Objective To compare low-density lipoprotein cholesterol （LDL-C） value measured by homogeneous assay and the value calculated by Friedewald formula, and to evaluate the correlation and the potential bias of these two methods for clinical application. Methods LDL-C, high-density lipoprotein cholesterol （HDL-C）, total cholesterol （TC） , triglycerides（TG） were measured from a total of I 180 fasting serum samples. The correlation coefficient and linear equation between direct measurement of LDL-C and calculation from Friedewald equation were obtained, and they were used to evaluate outcome bias between these two assays. Then the Friedewald method was compared with the homogeneous assay for its ability to appropriately classify patients into different gToups. The positive predictive value （PPV） and negative predictive value （NPV） at the medical decision cut-points were evaluated, using the LDL-C homogeneous assay as the reference. Results The correlation coefficients were 〉 97% （P 〈 0. 001 ）. All the outcome biases were within the upper limits and lower limits of the corresponding confidence interval at each medical decision cut-point. LDL- C concentrations determined by LDL-C Friedewald method correctly classify 〉 85% of the subjects into different cutpoints except those whose concentrations were between 3. 12 and 3.64 mmol/L by homogeneous assay. The PPV and NPV of the Friedewald method were 〉 98% and 〉 85% respectively at different medical cut-points despite a lower PPV （77. 8% ） at 3. 64 mmol/L. Conclusions There is good correlation between LDL-C homogeneous assay and Friedewald method. The outcome bias of Friedewald method is acceptable. The Friedewald equation has the ability of differentiate ＂positive＂ and ＂negative＂ suhjects at different medical out-points, except for the samples with concentrations between 3.12 -3.64 mmol/L. For those the Friedewald equation might give higher values than the homogeneous assay.
出处
《检验医学》
CAS
北大核心
2007年第3期315-319,共5页
Laboratory Medicine
