Posted by Sten Westgard, MS
I came across a troubling admission in an exchange of letters about a study of the quality of POC glucose meters:
"....In our study, the meters had a CV of 2.5-5%. A glucose level of 7.0 mmol/L could thus read anywhere between 6.3 and 7.7 mmol/L with 95% confidence. With this degree of variance, single readings using these meters should not be used for either the diagnosis of diabetes, or for making critical clinical decisions, e.g. initiating insulin. Glucose meters are best used by analysing the results of multiple readings over time, which reduces the error involved, rather than relying on single readings...."
Matthew Cohen, Jonathan Shaw, Response to the letter by Cheung et al., Diabetes Research and Clinical Practice 72 (2006) 218
Maybe I'm missing something, but wasn't the point of POC devices to get quick, accurate readings you can act upon? If the only way to get useful results out of POC meters is to make multiple readings over time, why not just order the lab test? You already have to wait - so why not get a good, accurate single result from the better method?
It's my naivete, I'm sure.
Comments