PERFORMANCE OF GLUCOSE MONITORS

Seven types of glucose monitors obtained from three manufacturers were evaluated using venous whole blood. In each case the coefficient of variation was below 5%. Glucose measurements obtained from the monitors were compared with those from a laboratory reference instrument, the Beckman Synchron CX3. This instrument uses blood plasma, and the values obtained were adjusted to whole blood to make them comparable to the values obtained with the other monitors. Three groups of tests, each covering a broad range of glucose values, yielded precise and comparable results. Three Accu-Chek IIM monitors gave similar results-all about 6% to 8% low. Three One Touch monitors gave results that varied depending on the instrument-9% low, 13% low, and 16% low. The results were 20% high using one Glucometer QA and 15% high using a Glucometer 3. In our tests, the new Accu-Chek III seemed to perform similar to the Accu-Chek IIM. The most accurate and precise results were obtained with two new monitors, the One Touch II and Accu-Chek Easy, both of which employed test strips that completely absorbed the drop of blood and did not require wiping to remove excess blood. The One Touch II was 4% low. The Accu-Chek Easy was 5% high at low glucose values and 2% low at high glucose values. This tendency to have high readings at low glucose values was exacerbated by exposing the strips to the atmosphere for several minutes (prior to contact with the blood sample) and/or by high temperatures. These problems plus a small number of inaccurate monitors prompted the manufacturer to recall the Accu-Chek Easy monitors and test strips. The manufacturer plans to reissue a no-wipe system. The possibility of patient contamination during the use of blood monitors and the factors related to the accuracy of blood glucose measurements are discussed.