User Verification of Precision and Estimation of Bias; Approved Guideline
|Publication Date:||1 September 2014|
This guideline was developed as a protocol for simultaneously verifying a manufacturer's claims for precision of a measurement procedure and the trueness of the measurement procedure relative to the assigned values of materials with known concentrations.
The precision verification section of the guideline was developed for situations in which the performance of the procedure has been previously established and documented by experimental protocols with larger scope and duration. It has relatively weak power to reject precision claims with statistical confidence, and should only be used to verify that the procedure is operating in accordance with the manufacturer's claims. This document is not intended to establish or validate the precision performance of a measurement procedure.
The bias estimation section of the guideline relies on 25 or more measurements by the candidate procedure, made over five or more days, to estimate the measurand concentrations of materials with known concentrations. These estimated measurand concentrations are compared to the assigned measurand concentrations of the materials to estimate bias. The observed bias is a measure of trueness if a high-quality measurement procedure was used to assign the concentrations of the materials.
Because this document's scope is limited to verification of precision and estimation of bias, other more rigorous CLSI protocols (eg, see CLSI documents EP06,3 EP17,4 and EP285) are employed to validate the measurement procedure's performance against the user's needs. CLSI documents EP051 and EP096 were developed to assist manufacturers in establishing the performance of a diagnostic device for precision and trueness, respectively. (Also, see CLSI documents EP06,3 EP17,4 EP28,5 and EP10.7) CLSI document EP107 is intended for the rapid preliminary evaluation of precision, bias, sample carryover, drift, and nonlinearity.
One may also note that the EP15 protocol has an implicit assumption: Namely, that if the estimated precision and bias are acceptable, then the overall error (eg, total analytical error) of the measurement procedure is acceptable. This implied model can lead to an underestimation of the total analytical error8 in cases in which other effects are important. Besides conducting more extensive evaluations mentioned above, one could also consider performing the protocol within CLSI document EP21.9 This protocol is a direct estimation of total analytical error, and does not rely on a model.