Related Experiment Videos
The use of portable blood glucose monitors by trained lay operators
E M Campbell1, S Redman, P R Dunkley
1Faculty of Medicine, University of Newcastle, Callaghan, NSW.
Objective:
To evaluate the use of Reflolux II blood glucose reflectance meters by trained lay operators in a community survey.
Design:
A random household survey involving assessment of participants' blood glucose levels by Reflolux II. For a random subsample of participants an additional sample of blood was placed onto filter paper strips and stored for later examination and assay.
Setting:
The survey was conducted in the Newcastle and Lake Macquarie areas of New South Wales.
Participants:
Of the 1275 participants 15 years and over, 1229 consented to have their blood glucose assessed and 163 of these also had blood stored on filter paper strips. Data were collected by 25 interviewers.
Main Outcome Measures:
Application of blood spots to filter paper assessed the interviewers' capacity to collect samples of minimal acceptability for Reflolux II operation. Samples were considered adequate if they covered at least a 6 mm soaked disk and there was no evidence that they had been smeared or applied from both sides of the paper. The Reflolux II readings obtained by the interviewers were compared with the blood glucose values from assay of the filter paper samples.
Results:
Only 63% of the 163 samples collected on filter paper were of acceptable quality for adequate Reflolux II operation. The overall correlation between Reflolux II blood glucose values and those determined from the blood collected on filter paper was good (r = 0.893). However, three of the eight interviewers who had five or more readings for comparison achieved correlation coefficients of < r = 0.20. Of the survey participants identified as having elevated blood glucose levels (> 8 mmol/L) by the filter paper assay, 58% (7 of 12) were not detected by the lay operated reflectance meters. Twenty-two per cent of the individuals assessed by the reflectance meter, compared with 0.5% of a population sample surveyed by the National Heart Foundation, had blood glucose values of less than 3.5 mmol/L.
Conclusion:
It would appear that there may be a considerable error rate in blood glucose values obtained by lay operators using reflectance meters, resulting in underestimation of blood glucose levels. The findings have implications for community awareness programs for diabetes and highlight the need for careful training and monitoring of lay operators in their obtaining of finger-prick blood samples and their use of reflectance meters.