Ward and laboratory assessment of two second-generation blood-glucose monitoring systems

We present an assessment of two blood-glucose monitoring systems--the Ames Glucostix reagent strips read on the Glucometer II and the Boehringer-Mannheim BM-Test-Glycemie 20-800 test strips read on the Reflolux II meter. Many nurses in the wards and a technologist in the laboratory assessed the systems. All operators achieved good precision with both systems at various glucose levels, the coefficients of variation for the technologist being less than 8% and those for the nurses being less than 11%. The Boehringer-Mannheim system gave good agreement with the laboratory values, with a mean bias of +0.14 mmol/L and limits of agreement between -1.38 mmol/L and +1.66 mmol/L. The Ames system gave poorer agreement, with a bias of +0.80 mmol/L and limits of agreement between -2.12 mmol/L and +3.72 mmol/L. Our study shows that both systems are capable of providing clinically-acceptable results when used by trained personnel. Although it took longer to use, the operators unanimously preferred the Boehringer-Mannheim system because generally it was easier to handle. This is due to a smaller amount of blood being needed which is easier to apply and to remove from the strip, and the cleaning material is not as likely to be left on the strip. Also, strip insertion is easier and there is sufficient time for this after the removal of blood.