http://jkrouwer.wordpress.com/2008/10/29/comments-on-%e2%80%9cmeasurement-uncert\
ainty-is-not-synonym-of-measurement-repeatability-or-measurement-reproducibility\
%e2%80%9d/Comments on "Measurement uncertainty is not synonym of
measurementrepeatability or measurement reproducibility"Of course, I agree with
this statement (1) and here are some comments.First, some definitions and
observations.1. A repeatability condition is defined as a "condition
ofmeasurement, out of a set of conditions that includes the same
measurementprocedure, same operators, same measuring system, same operating
conditionsand same location, and replicate measurements on the same or similar
objectsover a short period of time"2. A reproducibility condition is
defined as a "condition ofmeasurement, out of a set of conditions that includes
different locations,operators, measuring systems, and replicate measurements on
the same orsimilar objects"3. GUM includes two sources of uncertainty both
expressed as standarddeviations: type A, which is characterized by measurements
and type B, whichis characterized by either measurements or more commonly by
experience orassumptions.4. In the real world of commercial diagnostic
assays, reproducibilityis almost always larger than repeatability.5. It is
logical to assume that the reason for #4 is uncorrectedsystematic effects.6.
A reason that some effects are uncorrected (or not bettercorrected) is
economics.Now if one could take an infinite set of measurements for a
diagnosticassay, would there be a difference between reproducibility and
uncertaintyof measurement? I maintain the answer is no, they will be the same.
All ofthe type B effects (and type A) will be expressed in an infinite set
ofmeasurements.For a shorter set of measurements, reproducibility and
uncertainty ofmeasurement will be different, although for diagnostic assays, one
routinelyhas large sets of quality control data (reproducibility) that
spanrelatively long times.One problem with this quality control data is that it
is not patient dataand thus some effects cannot be sampled such as patient
interferences.(Postulating them through assumptions in GUM is not easy
either).Another consideration is the types of effects that manifest themselves,
suchas calibration bias and non calibratable reagent lot effects. In
principle,these systematic effects could be made smaller but aren't since
economicsprevent it. Since these effects can be relatively large, given a long
enoughsampling time, these effects and the other unknown effects which
areexpressed over time will approximate an uncertainty of measurement
approachalthough as stated above, the quality control results will never account
foreffects such as patient interferences.So although measurement uncertainty is
not synonym of measurementrepeatability or measurement reproducibility, a
reproducibility experiment,conducted over a long enough time will probably give
a result that issimilar to an uncertainty of measurement approach (save for the
patientinterference problem).And finally, it is a lot easier to calculate a
standard deviation on somequality control data, than to go through a proper
uncertainty of measurementprocedure.References1. Paul De Bièvre
Measurement uncertainty is not synonym ofmeasurement repeatability or
measurement reproducibility Accred Qual Assur(2008) 13:61-62This entry was
posted on Wednesday, October 29th, 2008 at 12:25 pm and isfiled under Clinical
laboratory statistics, GUM, ISO, laboratory medicine.You can follow any
responses to this entry through the RSS 2.0 feed. You canleave a response, or
trackback from your own site.