| Objective: To explore the precision and accuracy of four POCT glucose meters in clinical with three different analyzing principles(sugar glucose dehydrogenase electrochemical method, electrochemical processes, photochemical method of glucose oxidase) whether meet the national standard, compared with the automatic biochemical analyzer of Roche Cobas8000-c701 in laboratory, and to determine the differences and correlation of different glucose meters in testing peripheral whole blood glucose and venous serum glucose after fasting blood sugar and/or blood sugar load, in order to provide a reference for clinical and to better guide for clinical work.Methods:1 The inclusion criteria: 451 patients with diabetes or without diabetes needing venous blood test and oral 75 g anhydrous glucose powder(glucose tolerance test) were selected from our hospital outpatient or ward. Gender and age were unlimited.2 Test of precision: the precision of the four POCT glucose meters with three different analyzing principles was tested with their own quality control liquid respectively. The standard deviation(SD) and coefficient of variation(CV) were calculated referring to the United States released CLSI EP15- A2 file and set by the ministry of health in our country, the medical institutions of the portable blood glucose meter management and clinical operation standard(trial) 2010 "(hereinafter referred to as" specifications ") regulation: to control the quality when liquid glucose concentration is less than 5.5 the tendency for L, the SD should be 0.42 mmol/L or less; When more than 5.5 mmol/L, CV % should be 7.5% or less.3 Accuracy analysis: deviation of each POCT glucose meter measuring fasting and(or) random time points after blood sugar load was compared with biochemical analyzer respectively. Formula:(the value of finger tip blood glucose – the value of corresponding time point venous serum glucose determined by biochemical analyzer), with reference to the specification: when blood glucose concentration is less than 4.2mmol/L, at least 95% glucose results should be within the deviation range tendency for 0.83 mmol/L; When the blood glucose concentration is greater than or equal to 4.2mmol/L, at least 95% glucose results should be within ±20% deviation.4 The data collection: using two or three or four POCT glucose meters to measure fingertip blood sugar according to the voluntary degree of each patient’s. Meanwhile, the corresponding hand elbow venous blood was collected to determine serum glucose with automatic biochemical analyzer.5 Statistical analysis: The standard deviation(SD), coefficient of variation(CV), precision and deviation were calculated. The differences and correlation between different kinds of glucose meters and biochemical analyzer were analyzed. All data were analyzed by SPSS 13.0. If the distribution of data was normality, data was described with mean ± SD, and comparison between two groups was conducted with t test, in addition, correlation analysis was conducted with Pearson correlation test. If the distribution of data was not normality, data was described with median and QR, and comparison between two groups was conducted with Wilcoxon test, in addition, correlation analysis was conducted with Spearman correlation test.Results:1 The results of precision test Roche glucose meter, batch CV % in high concentration quality fluid was 2.4%, less than 3.3% manufacturer statement, CV % in different batches was 3.1%. In low concentration batch SD was 0.06 mmol/L and SD in different batches was 0.08 mmol/L. Abbott glucose meter, batch CV % in high concentration quality fluid was 3.0%, less than 3.2% manufacturer statement, CV % in different batches was 3.3%. In low concentration batch SD was 0.08 mmol/L and SD in different batches was 0.09 mmol/L. Yuwell glucose meter, in low concentration batch SD was 0.08 mmol/L and SD in different batches was 0.10 mmol/L. Turomo glucose meter, batch CV % in high concentration quality fluid was 3.6%, less than 3.8% manufacturer statement, CV % in different batches was 3.3%. In low concentration batch SD was 0.07 mmol/L and SD in different batches was 0.09 mmol/L. All the results were below the national "specification" in glucose meter repeatability(precision) requirements.2 The results of accuracy analysis Referring to the venous serum glucose results, the deviation analysis results of three kinds of analysis of the principle of 4 sets of POCT glucose meters: in fasting state, when blood sugar concentration <4.2mmol/L, the ranges of four glucose meters were-0.63~0.27mmol/L,-0.57~0.23mmol/L,-0.77~0.08mmol/L, ——, respectively; when blood sugar concentration ≥4.2mmol/L, the ranges of four glucose meters were-17.2%~15.9%,-18.4%~19.2%,-16.2%~18.3%,-17.4%~12.0%. In glucose load, when blood sugar concentration <4.2mmol/L, the ranges of four glucose meters were-0.59~0.18mmol/L,-0.71~0.52mmol/L,-0.47~0.51mmol/L and-0.78~0.37mmol/L, respectively; when blood sugar concentration ≥4.2mmol/L, the ranges of four glucose meters were-17.6%~18.1%,-17.2% ~ 18.5%,-13.7% ~ 16.9% and-18.7% ~ 12.9%. All the deviation ranges were lower than the national "specification" requirements.3 The comparison and correlation analysis between the blood glucose monitor and the biochemical analyzerIn the analysis of the Roche blood glucose monitor and the biochemical analyzer, P=0.827 >0.05, r=0.932 in the fasting state, and P=0.065 >0.05, r=0.969, no significant difference was observed.In the analysis of the Abbott blood glucose monitor and the biochemical analyzer, P<0.001, r=0.938 in the fasting state, and P<0.001, r=0.969, there were significant differences.In the analysis of the Yuwell blood glucose monitor and the biochemical analyzer, P=0.689 >0.05, r=0.966 in the fasting state, and P=0.122 >0.05, r=0.974, no significant difference was observed.In the analysis of the Terumo blood glucose monitor and the biochemical analyzer, P<0.001, r=0.958 in the fasting state, and P<0.001, r=0.973, there were significant differences.4 The overall comparison between the blood glucose monitor and the biochemical analyzer in different glucose levelsIn the analysis of the Roche blood glucose monitor and the biochemical analyzer, there were significant differences(all P<0.001), when blood sugar concentration <4.2mmol/L or ≥16.6mmol/L; when blood sugar concentration was ≥4.2~<6.7mmol/L, or ≥6.7~<11.1mmol/L, or ≥11.1~<16.6mmol/L, there was no significant difference.In the analysis of the Yuwell blood glucose monitor and the biochemical analyzer, there were significant differences(all P<0.001), when blood sugar concentration <4.2mmol/L, or ≥11.1~<16.6mmol/L, or ≥16.6mmol/L. when blood sugar concentration was ≥4.2~<6.7mmol/L, or ≥6.7~<11.1mmol/L, there was no significant difference.Conclusions:1 Three kinds of analysis of the principle of 4 sets of POCT glucose meter are accord with national blood glucose detection system for glucose meter precision and accuracy requirements.2 The results of Roche and Yuwell blood glucose monitors are similar to the biochemical analyzer in fasting state and glucose state. Compared with the biochemical analyzer, Abbott blood glucose monitor is much lower, while Terumo blood glucose monitor is higher.3 There are correlation between different glucose meters and biochemical analyzer after fasting blood sugar and/or blood sugar load; And the correlation is the best between Yuwell blood glucose monitor and biochemical analyzer.4 Compared with the biochemical analyzer, the results are not accurate when the blood glucose is much lower or higher in the analysis of the Roche or Yuwell blood glucose monitor. |
PDF Full Text Request