To judge how reliably patients perform capillary blood glucose testing over time with recurrent practice at home and to assess if a clinic glucose meter is an acceptable alternative to the clinical laboratory for monitoring patient performance.
We compared capillary blood glucose readings obtained by patients with their own equipment and the venous blood glucose determinations by the clinical laboratory at three biweekly visits during the initial phase in 40 subjects attending the diabetes clinic at the Veterans' Affairs Medical Center in Phoenix, Arizona. We also compared patient-generated readings using their own equipment and the readings obtained by the clinic glucose meter and strips at five weekly visits during the second phase in 11 subjects who continued further participation. Error grid analysis was used for both comparisons. Capillary blood glucose readings obtained with clinic glucose meter and strips on one hand were correlated with venous blood glucose levels determined by the clinical laboratory.
During the initial phase, 30 subjects consistently obtained clinically acceptable comparisons (zone A on the error grid, i.e., within 20% of the laboratory value) or improved over time, 9 subjects showed deterioration, and 1 subject failed to obtain zone A results on any of the visits. Three subjects who had consistently obtained zone A results during the initial phase maintained their performance, whereas eight subjects who had failed to achieve zone A values by the end of the initial phase gradually improved and ultimately achieved zone A values by the end of the study. A highly significant correlation was noted between clinic meter readings and laboratory values (r = 0.93, P < 0.00001).
Clinically acceptable user proficiency in capillary blood glucose testing can be maintained in most subjects, with recurrent intensive education during follow-up clinic visits. Therefore, we recommend that these comparisons be performed and patient's technique be observed at each visit to monitor their performance. The clinic glucose meter is a suitable alternative to a clinical laboratory for user proficiency checks.