Comparative Study Of The Blood Glucose Values That Venous Blood,arterial Blood And Capillary Blood In Critically Ill Patients

Objective: To compare the results reliability of the blood glucose values that arterial blood using a blood gas analyzer and capillary blood using glucometer in critically ill patients. And Analysis of clinical pathology and medicine factors. The aim is to investigate the reliability and influence factors of the two measurements. So as to provide references for evaluating the condition of patients and identifying the treatment plan of tight glucose control of insulin, with clinical application of tested blood glucose level in a faster and more convenient way.Methods: In the morning, comprehensive ICU patients aged ≥ 18 years and need to monitor venous blood glucose of central laboratory, arterial blood gas analyzer and capillary blood glucose at the same time, were enrolled in this study. three kinds of blood samples were drawn and tested simultaneously. The patients’ real-time laboratory parameters(hematocrit, PH value etc.) and assessment of critical degree(APACHEⅡ score) were recorded. Compared with reference standard of central laboratory, the agreement and the accuracy between arterial and capillary for blood glucose was analyzed.The influence factors of the arterial and capillary for blood glucose were analyzed statistic by the logistic regression method.Results: The study involves 170 cases.(1) Arterial blood glucose and capillary blood glucose showed a strong correlation with venous blood glucose(r = 0.927 and 0.861, P <0.001).(2)the mean difference(bias) was 0mmol/L for arterial blood glucose and venous blood glucose, showed good consistency. the mean difference for capillary blood glucose and venous blood glucose was 0.3mmol/L, the consistency not so good;(3) The average blood glucose level of capillary(8.16±3.09mmol/L) was higher than laboratory reference(7.83±3.59mmol/L, P<0.01),but there was no Statistical significance for arterial blood glucose measurement.(4) The rates of deviations for capillary blood measurements(31.1%)significantly higher than arterial blood glucose measurement(8.82%,P=0.001).(5) In the different group which grouped according to blood glucose level, arterial blood glucose and venous blood glucose, all good consistencies. During laboratory reference value <4.4mmol/L, the bias was 0.73mmol/L for capillary blood glucose and venous blood glucose, showed poor consistency. The capillary blood glucose methods have a high deviation rate and significantly higher than other group which different glucose levels(P<0.05), when glucose level is lower than 4.4mmol/L.(6) In the group of abnormal fingertip skin, the bias was 0.6mmol/L for capillary blood glucose and venous blood glucose, showed poor consistency and the rate of deviation also increases obviously than the normal fingertip group( P=0.000).(7)On the impact factors of accuracy for capillary blood glucose measurements, The patients’ glucose level, the application of vasopressor, hematocrit level, and APACHEⅡ scores were into the logistic regression equation, for arterial blood glucose measurements,Pa O2 and HCT value were into the logistic regression equation.Conclusion:(1)Arterial blood glucose measurements and capillary blood glucose measurements had a strong correlation with laboratory method, capillary blood glucose tended to yield higher real blood glucose values of the patients, Especially when blood glucose levels below 4.4mmol / L, the risk of inaccurate measurements by capillary blood method in critically ill patients was higher than arterial blood gas method. in this case, we should base on the arterial blood gas analysis values.(2) The accuracy of capillary blood measurements and its consistency with venous blood glucose was significantly effected by fingertip skin condition, blood glucose level, critical degree, application of vasopressor and hematocrit level.(3) The accuracy of arterial blood gas glucose measurements was significantly effected by Pa O2 and HCT level.