Accuracy of blood glucose concentrations determined by four visually read reagent strips

Two new reagent strips have recently been introduced for blood glucose measurement by direct visual reading. Results obtained with these strips (Glucostix and Hypogard GA) were compared with those obtained using other commonly employed strips (BM-Test-Glycemie 1-44 and Visidex II) and a standard laboratory method. Blood glucose estimations were performed on samples of venous blood drawn from 125 patients attending the diabetic clinic using each of the four strips and the laboratory method. Results obtained with the strips correlated with the laboratory values as follows: BM-Test-Glycemie 1-44, r=0.93; Glucostix r=0.93; Hypogard GA r=0.87 and Visidex II r=0.92. The lower correlation with Hypogard GA reflected consistent underestimation of the laboratory value (slope of regression line = 0.63). Readings in error by 20% or more were: BM-Test-Glycemie 1-44, 14%; Glucostix, 15%; Hypogard GA, 31%, and Visidex II, 14%. With Hypogard GA strips, 57% of readings above 16 mmol/l were inaccurate. We conclude that Hypogard GA strips cannot be recommended for direct visual reading. Acceptable results may, however, be obtained using the other three strips.Reagent strips allow reasonably accurate determinations of blood glucose concentrations when used with a reflectance meter.(1, 2) Nevertheless many diabetic patients prefer to read the reagent strips visually. This method avoids the problems associated with meter calibration, is cheaper and also is more portable. Direct visual readings with BM-Test-Glycemie 1-44 (Boehringer Corporation) and Visidex II (Ames) have been shown to be acceptable in the hands of medical and technical personnel.(3, 4) Recently two new reagent strips have been marketed, Hypogard GA (Hypogard UK Ltd) and Glucostix (Ames), and it is claimed that they are also suitable for direct visual reading. To test the validity of these claims, we have compared results obtained using the newer strips with readings from BM-Test-Glycemie 1-44 and Visidex II and with a standard laboratory method.