血脂检测结果一致性比较及其对临床诊断决策的影响

摘要

目的 尝试通过统一定标品,改善不同方法检测血脂项目的结果一致性,减少结果不一致可能对临床诊疗造成的影响.方法 方法学评价.随机收集2012年3月13至19日复旦大学附属中山医院住院患者(年龄37～68岁,男238例,女162例)血清标本400份,每份标本均使用5种试剂[德国罗氏诊断(Roche)、日本和光纯药(Wako)、日本日东纺医疗(Nittobo)、德国德赛诊断(Diasys)、日本积水医疗(Sekisui)]分别检测血清TC,TG,HDL-c,LDL-c,Apo A-Ⅰ及Apo B.使用自制定值血清对各检测方法进行校正,并重复检测上述标本.使用变异系数描述同一标本在校正前后不同检测方法间的检测结果差异.按TG及LDL-c水平分别将标本分为4组及5组,使用t检验比较各组间血脂项目在校正前后不同检测方法检测结果差异.以TC≥6.22 mmol/L及LDL-c≥4.14mmol/L作为判断标准,比较不同检测方法间检测结果差异对临床决策的影响.结果 血脂项目不同检测方法间的检测结果存在差异(CV均值为4.50％～10.46％),定值血清校正后差异减小(CV均值为2.82％ ～6.71％).根据LDL-c及TG水平将患者分组后,各组内检测结果差异明显(CV:LDL-c4.03％～ 12.14％；TG 4.02％ ～ 18.42％),CV差异程度与LDL-c水平无关,但随TG水平的增高而增大.除TG高值组外(＞5.65 mmol/L),其他各组检测差异均可通过定值血清校正加以改善(CV:LDL-c 2.35％ ～6.91％；TG 2.08％ ～8.91％).不同检测方法间检测结果的不一致会明显影响临床决策,差异可达19.83％ (TC)和34,48％ (LDL-c).结论 不同检测方法间的检测结果存在较为显著的差异.各血脂项目使用相应的同一份定值血清进行校正后,不同检测方法间的结果差异明显减小但是无法完全消除.%Objective To improve the harmonization of lipid results and reduce the impact on clinical practice by using the unified calibration approach.Methods Sera from 400 inpatients in Zhongshan Hospital of Fudan University (March 13 2012 to March 19; age:37-68; 238 male; 162 female) were collected.Serum TC,TG,HDL-c,LDL-c,Apo A-Ⅰ and Apo B levels were measured by 5 different methods including Roche,Wako,Nittobo,DiaSys and Seikisui.Roche manufacturers were used as the reference method,and fresh samples were chosen as calibration serum to calibrate the five methods.The values of the calibration sera were similar to those of Roche calibrators.After being calibrated by calibration serum,we retested all samples and compared the results before and after calibration.Variation coefficient was used to describe the difference among the results of same sample that measured by each method before and after calibration.According to the concentrations of TG or LDL-c,we classified the samples into 4 or 5 groups and compared the results of each group,including different systems as well as before and after calibration.When TC values ≥6.22 mmol/L or LDL-c values ≥4.14 mmol/L was taken as the assessment criterion,we compared the influence of results from different systems on clinical treatment by calculating numbers of patients who need clinical treatment.Results The data showed that the harmonization was not well in different detection systems (CV 4.50％-10.46％),however,CV values could be reduced (2.82％-6.71％) after being calibrated by the calibration serum.There were significant difference among the groups (CV:LDL-c groups 4.03％-12.14％ ;TG groups 4.02％-18.42％),the CV values had no correlation with the LDL-c levels,while they raised with the increase of TG levels.In addition to the group that TG values ＞5.65 mmol/l,the CV values in other groups could be improved by calibration serum (CV:LDL-c groups 2.35％-6.91％ ;TG groups 2.08％-8.91％).Furthermore,the harmonization of the lipid and lipoprotein results would influence the clinical treatment.Conclusions The lipid and lipoprotein results of different detection systems were still significantly different.This may have an impact on clinical diagnosis and treatment.Using the same calibration serum to calibrate all systems could reduce the difference of results significantly and improve the harmonization of lipid as well as lipoprotein detection.TG high value had an impact on lipid and lipoprotein detection and could not be eliminated by calibration.