[Accuracy of blood glucose measured in the hospital]

Glucometer determination of capillary blood sugar (CBS) has replaced determination of glucose in the urine and in venous blood for monitoring and treating inpatient diabetics. CBS determinations were performed by 49 ward nurses chosen at random. Each performed 2 CBS measurements using an Ames Glucometer; glucose levels were also determined using a glucose-analyzer in blood obtained from the same finger puncture. Accuracy of technique was evaluated on a scale of 1 (inaccurate) to 5 (accurate) by the hospital diabetes nurse and a questionnaire was completed. A deviation of < 20% between Glucometer and glucose-analyzer results was considered satisfactory. 58% of nurses deviated by 20% or more. There was no significant correlation between accuracy of measurement and the nurse's age, seniority, department, source of instruction or number of weekly determinations. 71% of non-registered nurses and 45% of registered nurses tested inaccurately. There was a significant correlation between performance score given by the diabetes nurse and accuracy of measurement. We therefore conclude that the inaccuracy of Glucometer measurements by ward nurses is significant, registered nurses need to be trained and certified periodically in Glucometer use, an ongoing quality assurance program must be instituted on the wards, and "user-friendly" Glucometers are not suitable for ward use.