Fedele D, Corsi A, Noacco C, Prisco F, Squatrito S, Torre E, Iafusco D, Errico M K, Toniato R, Nicolucci A, Franciosi M, De Berardis G, Neri L
Diabetes Centre, Padova, Italy.
Diabetes Technol Ther. 2003;5(6):983-9. doi: 10.1089/152091503322641033.
The aim of this study was to compare glucose measurements between fingertip and forearm using the blood glucose (BG) monitoring system One Touch Ultra (LifeScan), an electrochemical sensor that requires only a very small drop of blood (1 microL). Patients with type 1 or type 2 diabetes were identified in five outpatient diabetes clinics. Participants were requested to use the One Touch Ultra at home for 1 week for the measurement of BG levels from both sites. Patients filled in a questionnaire about their experience with testing blood samples from fingertip and forearm. The agreement between the measurements from the two sites was assessed using linear regression analysis, mean absolute relative error (MARE), the Bland-Altman method, and Error Grid Analysis (EGA). Overall, 112 patients were recruited, of whom 58% had type 1 diabetes. Linear regression analysis showed an intercept of 17.7, statistically different from 0 (p<0.0001). The slope was 0.956, and the Pearson correlation coefficient was 0.95. A MARE of 12.1% (SD=11.8%) was obtained, with a greater deviation of the forearm values from the fingertip ones in the hypoglycemic range (MARE=22.3%; SD=21.7%). The Bland-Altman bias plot showed a mean bias of 10.2 mg/dL (SD=23.1), with no correlation between mean difference and average BG levels (r=0.02). The EGA showed that 89.2% of the values fell in zone A, 10.4% in zone B, and 0.4% in zone C. The vast majority of patients (71%) declared that the collection of blood from the forearm caused no pain or less pain than the traditional site. Only 17% of the patients declared that it was impossible to obtain any blood from the forearm, while 63% reported with satisfaction that the quantity requested was small. At the end of the study period, 32% of the participants indicated the forearm as the preferred test site. Alternative site testing on the arm, with a BG meter that requires only a very small drop of blood, is feasible and reliable under routine clinical conditions. When testing with the express purpose of detecting hypoglycemia, the finger still remains the recommended test site.
本研究旨在使用仅需极少量血液（1微升）的电化学传感器——“稳豪倍优”血糖仪（LifeScan），比较指尖和前臂的血糖测量值。在五家门诊糖尿病诊所中识别出1型或2型糖尿病患者。要求参与者在家中使用“稳豪倍优”血糖仪，为期1周，测量两个部位的血糖水平。患者填写了一份关于他们从指尖和前臂采集血样经历的问卷。使用线性回归分析、平均绝对相对误差（MARE）、布兰德-奥特曼法和误差网格分析（EGA）评估两个部位测量值之间的一致性。总体而言，招募了112名患者，其中58%患有1型糖尿病。线性回归分析显示截距为17.7，与0有统计学差异（p<0.0001）。斜率为0.956，皮尔逊相关系数为0.95。获得的MARE为12.1%（标准差=11.8%），在前臂值与指尖值在低血糖范围内的偏差更大（MARE=22.3%；标准差=21.7%）。布兰德-奥特曼偏差图显示平均偏差为10.2毫克/分升（标准差=23.1），平均差异与平均血糖水平之间无相关性（r=0.02）。EGA显示，89.2%的值落在A区，10.4%落在B区，0.4%落在C区。绝大多数患者（71%）宣称从前臂采集血液不会引起疼痛或比传统部位疼痛更轻。只有17%的患者宣称无法从前臂采集到任何血液，而63%的患者满意地报告所需血量很少。在研究期结束时，32%的参与者表示前臂是首选的检测部位。使用仅需极少量血液的血糖仪在手臂上进行替代部位检测，在常规临床条件下是可行且可靠的。当以检测低血糖为明确目的进行检测时，手指仍然是推荐的检测部位。
