Additional key phrases: blood glucose: Advantage meterBlood glucose meters have been in use in the North Down and Ards Hospital Trust for the last 8 years. Our initial assessments'< had shown the Boehringer Reflolux 11 to be the instrument of choice. Since then the Reflolux 11 has been replaced by the Boehringer Accutrend. Although this instrument was found to be reliable when used correctly.l in practice we have found problems due to insufficient sample being applied to the strip or blood collecting on the inside of the instrument. The Advantage glucose meter appears to overcome these difficulties.
METHODSThe Advantage glucose meter is a small hand held battery powered meter using dry chemistry reagent strips containing electrodes. Each vial of strips is coded and contains a calibration key which is snapped into place in the back of the meter before use. The number which then flashes on the instrument corresponds to the number on the side of the vial. The strip is inserted into the meter leaving a yellow coloured test window exposed. A minimum 9 J.LL blood sample applied to this area ensures no yellow is visible. The reaction is based on the oxidation of blood glucose by glucose dehydrogenase linked to reduction of ferri-to ferrocyanide. Electrons generated in this reaction pass down the electrodes and are measured in the instrument. The read time is 40 s and the stated range is 0·56-33·3mmol/L. The manufacturers state that results are not affected by haematocrit in the range 20--65% at a glucose level < 11·1 mmol/L and 25-55% at glucose levels> 11·1 mmol/L or by bilirubin up to a level of 342 J.Lmol/L. The instrument can operate capillary, venous fluoride oxalate, heparin but not ethylene-diaminetetraacetic acid (EDTA) samples.Precision was determined in the laboratory by replicate analysis of two venous, fluoride oxalate blood samples with medium and high levels of The effect of haematocrit was studied using two fluoride oxalate samples with blood glucose levels of 5·9 and 21·1 mmol/L, These were each centrifuged to separate plasma from cells and then recombined in different proportions to give blood samples with a haematocrit range of 0·09--64 and 0·09-D·58, respectively, on each of which blood glucose was measured in duplicate.Minimum sample volume was studied using decreasing amounts (9-2 J.LL) of fluoride oxalate blood samples with medium and high levels of blood glucose assayed in duplicate to determine at what level an error message appeared.Interference from bilirubin was studied using a reconstituted bilirubin standard (Sigma Chemical Co Ltd, Poole, Dorset BHI7 7BR, UK).
RESULTSThe within batch coefficients of variation at the blood glucose levels of 0,98, 1'9, 6·6 and 21·1 mmol/L were 18-4 (n= 12), 8·1 (n= 12), 3·5 (n=20) and 4·1% (n = 20), respectively.A difference plot! of the Apec and Advantage is given in Fig. 1. The mean (95% confidence interval) is -0·66 (-0'24 to -1'08)mmol/L, respectively, and limits and agreement are -3·8 to + 2·5 mmol/L, Mean blood glucose varied between 5·5-6·4mmol/L in t...