Biomedical Engineering Reference
In-Depth Information
to be utilized are defined in the protocol. Any planned deviation must be documented
in protocol or amendment
a priori
and approved by PI or SD and Laboratory
Director (LD).
5.1
General Guidance and Requirement
5.1.1
Matrices
An assay should be validated using the same biological matrix as the study samples.
If a matrix has limited availability, a suitable substitute may be used (e.g., plasma
ultrafiltrate to substitute cerebrospinal fluid) for the preparation of standard (STD)
calibrators. When spiking solution is needed to prepare STDs and Quality controls
(QCs) in matrix, the spiking volume should be less than 5 % of the total volume of
the matrix.
5.1.2
Stock Solution
Two individually prepared and verified stock solutions are used to prepare STDs
and QCs, respectively. Some laboratories require that the stability of the stock solu-
tion has to be established prior to formal validation. But some requires establishing
at least one time point during validation to cover the time of solutions used during
the validation.
5.1.3
Regression Model
The simplest model that adequately describes the concentration-response relationship
should be used, e.g., a linear model is simpler than a quadratic model. At the comple-
tion of the validation, evaluation of different regression models must be performed.
Justification for using a quadratic regression equation must be documented.
5.1.4
STD and QC Concentrations
Typically, six to eight different nonzero concentrations that cover the dynamic range
of the assay should be used to define the curve. Two STDs should be at concentra-
tions lower than the Low QC sample and two STDs should be at concentrations
above the High QC sample. Five levels of analytical QC, i.e., LLOQ (Lower Limit
of Quantitation), Low, geometric Medium (optional), Medium and High QC, are
recommended. In general, the QC concentrations should differ from standard con-
centrations. The corresponding concentration distributions are: LLOQ = same con-
