22台血糖仪与生化分析仪测定血糖的比对

目的对血糖仪与生化分析仪测定血糖进行比对，评估二者测定血糖的偏倚，以生化分析仪测定结果为参考对血糖仪进行校准。方法每批选择氟化钠一草酸钾抗凝血液样本5份（血糖浓度范围1.0-22.0mmol／L），每台血糖仪测定一批样本全血葡萄糖，生化分析仪测定该批样本血浆葡萄糖浓度，22台血液糖仪测定22批血液葡萄糖浓度，获得110个全血葡萄糖结果和110个血浆葡萄糖结果。结果22台血糖仪中有12台测定结果与生化分析仪测定结果的差异无统计学意义（P〉0.05），60个结果中，有8个结果的偏倚〉20％，最大偏倚为27.2％，平均偏倚为12.5％，有5个结果的偏倚〉0.83mmol／L。另外10台血糖仪与生化分析仪测定结果的差异有统计学意义（P〈0.05）。22台血糖仪测定的110个血糖结果中，≤4.2mmol／L的结果为23个，其中6个（26％）结果与生化分析仪测定结果偏倚〉0.83mmol／L，〉4.2mmol／L的结果有87个，其中21个（24％）结果与生化分析仪测定结果偏倚〉20％。以NCCLS应用准则为允许偏倚。22台血糖仪在5个不同葡萄糖浓度范围内偏倚部达到允许偏倚范围内的只有6台（27.3％），超过70％的血糖仪未达到NCCLS应用准则规定的允许偏倚范围。依据CLIA’88最大精密度目标，有15台血糖仪的最大精密度目标是可接受的。结论血糖仪与临床实验室生化分析仪测定的葡萄糖浓度之间存在较大的偏倚。如果做好分析前的质量控制、定期对血糖仪进行比对和校准，那么，对于所有采用胰岛素治疗的患者，特别是1型糖尿病患者，在血糖自我监测、检查无症状低血糖症等方面，血糖仪是不失为方便和可靠的仪器。

The comparing of 22 portable glucose meters and clinical chemistry analyzer for measuring blood glucose

Objective To assess the bias of measuring blood glucose between the portable glucose meter and the clinial chemistry analyzer ,to calibrate the portable glucose meters with the linear regression equation. Methods To collect 5 blood samples anticoagulated with sodium fluoride and potassium oxalate, the range of glucose concentration is 1.0-22.00mmol/L. Every portable glucose meter measured the glucose concentration of whole blood of 5 blood samples and clinical chemistry analyzer measured the glucose concentration of the plasma of 5 blood samples. These portable glucose meters were performed by nurses. The 110 whole blood glucose results and plasma glucose results were obtained respectively. Results There are no significant deference（P〉0.05） between the glucose results measured by 12 portable glucose meters and by clinical chemistry analyzer. But,in these 60 results, the biases of 8 results are higher than 20% ,the maximum is 27.2%, the average is 12.5%. The biases of 5 resuits are higher than 0.83mmol/L. There are significant deference （P〈0.05） between the glucose results measured by other 10 portable glucose meters and by clinical chemistry analyzer. In results of these 110 results, the results of ≤4.2mmol/L are 23, the biases of 6 （26%） results are higher than 0.83mmol/L. The results of 〉4.2mmol/L are 87, the biases of 21（24%） results are higher than 20%. In these 22 portable glucose meters, there are only 6（27.3%） that met the criteria oftbe NCCLS in the range of 5 glucose concentrations, that means, higher than 70% portable glucose meters could not met the criteria of the NCCLS. According the criteria of CLIA 88 ,the maximum precision of 15 portable glucose meters are acceptable. Conclusion There are larger biases of measuring blood glucose between the portable glucose meter and the clinical chemistry analyzer. Different assay methods and architecture lead to lack of correlation among meters , even from a single manufacturer. In fact , two meters of the same brand have been observe