Accuracy of four methods of home blood glucose monitoring in hemodialysis patients.

The accuracy of four techniques of home blood glucose monitoring in hemodialysis patients was compared. Arterial-venous blood samples were collected from 23 patients undergoing hemodialysis and analyzed for serum glucose concentrations using an automated laboratory method. Capillary-blood glucose concentrations were also determined in each patient by visual readings of the Chemstrip bG and Dextrostix reagent strips and by reflectometer readings of the StatTek Glucose and Dextrostix reagent strips. Readings of glucose concentration obtained by each of the four techniques were compared with laboratory-determined values, and the true and absolute deviations in these values for each technique were recorded. The effect of patient hematocrit on deviations in glucose concentration readings was evaluated by analysis of variance. The mean glucose concentration readings obtained using the visual and reflectometer (Dextrometer) readings of Dextrostix were substantially less accurate than the other two techniques when compared with serum glucose concentrations determined by the laboratory. The StatTek technique gave the smallest mean absolute deviation from laboratory-determined values. No correlation was found between the patient's hematocrit and either the true or absolute deviations in glucose concentrations. The Chemstrip bG and StatTek techniques are reasonably accurate methods for home monitoring of blood glucose concentrations in diabetic patients undergoing hemodialysis.