Dear Mr Kallner,
Thank you for the supporting information regarding the ACB measurement
verification, I have not previously seen this document and will probably test
the approaches sited at a later date.
I wanted to expand slightly on your statement regarding the useage of
a "coverage factor (usually 2)" to take into account distribution tail ends (I
presume this is the purpose?).
A factor of 2 takes into account the majority of the tail, but does not account
for the ability of the QC rules adopted for the analyser/test. Without including
this, one must assume that an analyser at any given time is totally without
error (random AND systematic), and that if an error of any magnitude were to
occur, it would be discovered instantly. This is of course unrealistic.
In real, practical, terms, the QC rules we use in the lab (e.g. Westgard rules; 1
result out of 2.5SDs, 10 on one side of the mean, etc) only allow us to
determine error to a specific degree. The more scrutinising the rule, the better
chance of discovering an error; running one QC is less powerful than running
10 QCs, also a QC rule of +/-2.5SD of the mean is more powerful than +/- 7SD.
As such, any calculation of the uncertainty in a result should not only include
a tail-end coverage factor as you describe, but also the ammount of error (in
standard deviations) detectable by a QC rule at a given statistical power (i.e.
0.9 for 90% of errors detected). This error can be calculated from the Power
Function graphs of these rules.
We rely (almost) entirely on the results of these few daily QC measurements
to tell us that our analysers are free from error. We can not just assume this
in calculation of uncertainty.
Best regards,
Andy Minett
------ACB discussion List Information--------
This is an open discussion list for the academic and clinical community working in clinical biochemistry.
Please note, archived messages are public and can be viewed via the internet. Views expressed are those of the individual and they are responsible for all message content.
ACB Web Site
http://www.acb.org.uk
Green Laboratories Work
http://www.laboratorymedicine.nhs.uk
List Archives
http://www.jiscmail.ac.uk/lists/ACB-CLIN-CHEM-GEN.html
List Instructions (How to leave etc.)
http://www.jiscmail.ac.uk/
|