Multicentric evaluation of eight glucose and four ketone blood meters

High precision meters for blood glycemia are mandatory for monitoring glucose status in patients, avoiding both hypo- and hyper-glycemia. Health care providers routinely used in both out- and in-patients point-of-care measurements of glucose and ketone. These measurements, frequently used for medical decisions, are known to be less accurate than those performed in laboratories. Our aim was to evaluate, within the frame of an Assistance Publique-Hôpitaux de Paris (AP-HP) multicentric study, the performances of eight glucose and four ketone meters, either connected or non-connected to a laboratory software.

Glucose meter accuracy, precision, correlation with plasma glucose determined in central laboratories and hematocrit interferences were determined according to the ISO 15197:2003 norm. The same norm was applied for the determination of accuracy, precision and recovery of ketone meters for B-hydroxybutyrate measurements.

Among those meters, seven were considered as acceptable for glucose measurement and two for ketone measurement. Since all meters do not fit clinically relevant criteria, meters' performances have to be evaluated before use in clinical practice.