SKML-Quality Mark for point-of-care test (POCT) glucose meters and glucose meters for home-use.

BACKGROUND Point-of-care glucose meters are used increasingly in semi- and non-professional context. The quality of glucose measurements depends on the quality of the equipment, the quality of use, and the pre-analytical conditions. In this article, a External Quality Assessment Scheme (SKML)-Quality Mark for point-of-care test (POCT) and self-test glucose meters is proposed, assessing analytical quality and technical quality. The analytical requirements are based on the biological variation concept, and a system to assess meters for the SKML-Quality Mark is described. Using the proposed system as an example, 14 meters were tested. METHODS The analytical quality of the POCT and self-test equipment was assessed for plasma calibrated glucose values by comparison with a trueness verified method traceable to the IFCC reference method in an accredited clinical laboratory. The concept is based on the biological variation system. The SKML-Quality Mark comprises the following criteria for blood glucose equipment: 1) Fulfilment of compliance with ISO 15197 and/or TNO guideline criterion; 2) Fulfilment of the total allowable error (TAE) criterion; 3) Fulfilment of the total allowable linearity bias criterion; 4) Fulfilment of the total allowable interfering substances bias criterion; and 5) Fulfilment of the haematocrit criterion. RESULTS The proposed SKML-Quality Mark system was tested on 14 commercial home-use meters. The TAE criterion is violated by two meters. The main reason for the violation is bias. For the majority of meters, the Passing and Bablok regression confidence interval does not include the intercept of 0.0 and slope of 1.0. In addition, Syx indicates dispersion around the line or non-linearity. The bias and total error at three different concentrations were investigated as part of the quality mark, resulting in disapproval of the Dicomed Sensocard Plus meter. The bias was significant for the Wellion Linus. With respect to interfering substances, bias of the same magnitude and sign as the bias without additive was seen for all meters for acetaminophen, indicating no additional interference. For ascorbic acid, an additional bias was seen for several meters. However, significant bias was demonstrated for the Sensocard Plus and Glucocard X-meter. CONCLUSIONS The biological variation concept offers a scientific basis for assessment of acceptable deviation. The concept is extended in the SKML-Quality Mark correcting for the limited number of measurements that can be performed while assessing home-use or POCT meters. The results show that three out of 14 meters fail the proposed quality mark.