5.11.4 If the result of
the microbial analyses is outside the specified acceptance criteria in Appendix
A, Table 2, the cannabis testing entity shall determine the cause and take
steps to remedy the problem until the result is within the specified acceptance
criteria.
Microbiology
Culture Methods - Qualitative and Quantitative
The quality control (QC) samples that are required for
culturing of cannabis and cannabis products using qualitative and quantitative
methods are included in Appendix A, Table 2.
Molecular Assays/Methods
The QC samples that are required for molecular (i.e.,
polymerase chain reaction (PCR), gel electrophoresis and probe-based qPCR with
or without melting curve analyses) analysis of cannabis and cannabis products
are listed in Appendix A, Table 3.
PCR positive DNA controls are used to verify that the PCR
master mix and reagents were prepared correctly to produce amplification of the
target nucleic acid. This type of positive control is analyzed with each PCR
run.
A PCR run is defined as a group of samples that are analyzed
at the same time under the same amplification conditions, using the same PCR
master mix, and in the same thermocycler. A PCR run may contain more than one
extracted sample batches.
A PCR run with multiple assays must have a DNA positive
control for each assay.
Inhibition controls are used to verify that interfering
constituents from a cannabis form, which may be carried over during isolation
of nucleic acids or organisms during sample processing, do not inhibit the PCR.
Because cannabis forms are constantly changing, inhibition positive controls
must be performed in every extracted sample.
PCR DNA negative controls are used to verify that the PCR
master mix and reagents were prepared correctly to produce amplification of the
target nucleic acid. This type of negative control is analyzed with each PCR
run. A PCR run is defined as a group of samples that are analyzed at the same
time under the same amplification conditions, using the same PCR master mix,
and in the same thermocycler. A PCR run may contain more than one extracted
sample batches.
A PCR run with multiple assays must have a DNA negative
control for each assay to verify that the amplification conditions are working
properly.
No template controls are used to verify no contaminating
nucleic acid has been introduced into the master mix. These controls are
prepared when template is added to the master mix. They are prepared as
separate PCR reactions to which aliquots of molecular-grade water or buffer are
added to the master mix in place of target nucleic acid or sample. A negative
result with this control indicates that the master mix and final processing
reagents are not contaminated. This type of negative control is analyzed with
each PCR run. A PCR run is defined as a group of samples that are analyzed at
the same time under the same amplification conditions, using the same PCR
master mix, and in the same thermocycler. A PCR run may contain more than one
extracted sample batch. A PCR run with multiple assays must have not template
controls for each assay to verify that the sterility of the assays.
One duplicate sample is required per run. A duplicate sample
is subjected to all of the same steps as the original sample. For qualitative
analyses, if the duplicate sample does not equal the sample result, the sample
and its duplicate must be reanalyzed. Consideration should also be given to
possibility of re-preparing and reanalyzing all associated samples. For
quantitative analyses, if the RPD of the sample and duplicate is greater than
100, the parent sample and duplicate sample must be reanalyzed. Consideration
should also be given to possibility of re-preparing and reanalyzing all
associated samples. When data are accepted, the result for the sample portion
designated as the "original sample" is reported.
5.11.5 Chemistry - Analytical, Organic and
Inorganic (Metals).
Quality control must be performed for each analytical,
organic and metal chemistry method.
Each cannabis testing entity shall maintain sufficient raw
data records to ensure the QC was performed at the frequency specified.
'Bracketing' of QC samples, rotating from across the
calibration curve range, is required.
QC samples must follow the first twenty (20) samples after an
initial calibration, every twenty (20) samples thereafter, and at the end of
testing samples. This would also apply to a continuing calibration.
Initial Calibration
A. Samples results must be associated with an
acceptable initial calibration. If the initial calibration is not acceptable,
corrective actions must be performed and all associated samples
re-analyzed.
B. No sample results
are to be reported nor data qualified for a failed initial
calibration.
C. Samples must be
analyzed under an initial calibration that was performed no more than one month
prior.
D. The following items are
required elements of an initial calibration:
1. The details of the initial calibration
procedures including calculations, integrations, acceptance criteria, and
associated statistics must be included or referenced in the method SOP. When
initial calibration procedures are referenced in the method SOP, then the
referenced material must be retained by the cannabis testing entity and be
available for review;
2. Sufficient
raw data records must be retained to permit reconstruction of the initial
calibration (e.g., calibration date, method, instrument, analysis date, each
analyte name, and analyst or technician's initials or signature; concentration
and response, calibration curve or response factor; or unique equation or
coefficient used to reduce instrument responses to concentration);
3. The cannabis testing entity must use the
most recent initial calibration analyzed prior to the analytical
batch;
4. Standards used for
calibration must be traceable to an international or national standard, when
commercially available; and
5. The
cannabis testing entity must have a written procedure addressing removal and
replacement of calibration standards.
E. The lowest calibration standard must be at
or below the lowest concentration for which quantitative data are to be
reported without qualification.
F.
The highest calibration standard shall be at or above the highest concentration
for quantitative data are to be reported without qualification.
G. Sample results must be quantitated from
the initial calibration and may not be quantitated from any continuing
calibration verification.
H.
Criteria for the acceptance of an initial calibration must be established
including any calculations (e.g., relative error, relative standard deviation).
1. R2 >= 0.990, and
2. Curve recovery of ±20% (and
±30% for the lowest point) for all points must be maintained.
I. The cannabis testing entity
must use and document a measure of relative error in the calibration as
specified in the method SOP.
Initial Calibration Verification
A. All initial calibrations must be verified
with a standard obtained from a second manufacturer or a separate lot prepared
independently by the same manufacturer.
B. Initial calibration verification is
performed by analyzing a test solution of known analyte concentration(s) after
initial calibration and prior to sample analysis.
C. In general, the check must be ± 20%
(± 30% for the lowest point) of the known value. Some individual methods
may require tighter tolerances (±10% of the known value).
Continuing Calibration Verification
A. The validity of the initial calibration
must be verified prior to sample analyses by a continuing calibration
verification with each analytical batch.
B. A CCV is performed by analyzing a test
solution of known analyte concentration(s) prior to sample testing on each
testing day and continued periodically during the analytical batch run, no less
frequently than once after each set of 20 samples, and at the end of each
run.
C. The CCV must be a standard
that is from the same vendor/lot that is used for the calibration
curve.
D. In general, the check
must be ± 20% (and ± 30% for the lowest point) of the known
value.
E. Calibration must be
verified for each compound, element, or other discrete chemical analyte, except
for multi-component analytes where a representative chemical, related substance
or mixture can be used.
F.
Instrument continuing calibration verification must be performed at the
beginning and end of each analytical batch, and at the frequency defined in the
method.
G. Sufficient raw data
records must be retained to permit reconstruction of the continuing instrument
calibration verification (e.g., method, instrument, analysis date, each analyte
name, concentration and response, calibration curve or response factor, or
unique equations or coefficients used to convert instrument responses into
concentrations).
H. Continuing
calibration verification records must explicitly connect the continuing
calibration verification data to the initial calibration.
I. If the continuing instrument calibration
verification results obtained are outside the established acceptance criteria,
the following steps must be taken:
1. If a
cause for the calibration verification failure is identified that impacts only
the calibration verification sample (e.g., a missed autosampler injection),
then analysis may proceed if a second calibration verification sample is
analyzed immediately and the result is within acceptance criteria. Samples
analyzed previously must be considered valid if bracketed by a passing
calibration verification sample. The cause for the failure of the first
calibration verification result must be documented; and
2. If the cause for the calibration
verification failure is not identifiable or has impacted other samples, then
corrective action must be performed and documented. Prior to analyzing samples,
the cannabis testing entity must demonstrate acceptable performance after
corrective action with calibration verification or a new initial calibration
must be performed. Samples analyzed prior to the calibration verification
failure must be reanalyzed.
J. Data associated with an unacceptable
calibration verification must not be reported with a qualifier. Qualifying the
data is not an acceptable approach.
Low Level Continuing Calibration
Verification
A. A LLCCV will
be run at the end of each analytical batch.
B. The measured value must be within ±
30 % of the prepared value.
C. The
cannabis testing facility entity shall prepare and analyze at least one of each
of the following QC samples for each analytical batch:
1. Negative Control, Method Blank;
and
2. Positive Control, Laboratory
Control Sample (LCS);
3. Matrix
spike sample;
4. Duplicate matrix
spike sample; and
5. Duplicate
sample.
D. The required
QC is summarized in Appendix 2, Table 4.
Negative Control - Method Blank (MB)
A. A method blank must be analyzed at a
minimum of one (1) per preparation batch.
B. The MB must be processed along with and
under the same conditions as the associated samples to include all steps of the
preparation and analytical procedure.
C. The MB is used to assess the samples in
the preparation batch for possible contamination during the preparation and
processing steps.
D. The measured
concentration of each analyte in the MB or LRB must be < LOQ or
MRL.
E. Procedures must be in place
to determine if a MB or LRB is contaminated. While the goal is to have no
detectable contaminants, each method blank must be critically evaluated as to
the nature of the interference and the effect on the analysis of each sample
within the batch.
F. The source of
contamination must be investigated and measures taken to minimize or eliminate
the problem and affected samples reprocessed if the concentration of a targeted
analyte in the blank is at or above the LOQ, if the blank contamination
otherwise affects the sample results as per the method requirements or the
individual project data quality objectives, and a blank is determined to be
contaminated. Samples associated with a contaminated blank must be evaluated as
to the best corrective action for the samples (e.g., reprocessing or data
qualifying codes). In all cases, the corrective action must be
documented.
G. Any affected samples
associated with a contaminated MB or LRB must be reprocessed for
analysis.
Positive Control - Laboratory Control Sample
(LCS)
A. The LCS is used to
evaluate the performance of the total analytical system, including all
preparation and analysis steps.
B.
The LCS must be carried through the entire sample preparation process and
analyzed.
C. The LCS must be spiked
with all target analytes at a mid-level concentration in the curve.
D. The LCS must be analyzed at a minimum of
one (1) per preparation batch.
E.
The LCS is a quality system matrix, known to be free of analytes of interest,
spiked with known concentrations of analytes that are within the calibration
range.
1. A laboratory control sample (LCS)
may be used in place of a continuing calibration verification (CCV) (but not as
a replacement for a failing CCV) for methods where the calibration goes through
the same process as the LCS. Note that the more stringent acceptance criteria
must be met.
2. The matrix spike
may be used in place of this control as long as the acceptance criteria are as
stringent as for the LCS.
3. The
lab may use commercially available or pre-prepared standards (separate from
calibrators) for QC.
F.
All analyte concentrations must be within the calibration range of the
methods.
G. The individual LCS must
be compared to the acceptance criteria stated in the standard operating
procedure. The results of the individual batch LCS are calculated in percent
recovery or other appropriate statistical technique that allows comparison to
established acceptance criteria. The cannabis testing entity must document the
calculation.
H. When the acceptance
criteria for the positive control are exceeded, those sample results must be
investigated, and a corrective action implemented.
Matrix Spikes and Matrix Spike
Duplicates
A. Analyze an
actual sample with a known amount of standard added (matrix spike. MS). A
second portion of the actual sample used to prepare the MS that is spiked and
processed in the same manner as the MS (matrix spike duplicate, MSD).
1. For potency testing, a "representative
matrix" may be used to prepare the MS/MSD.
2. MS/ MSD shall be spiked at a midlevel
concentration with the target analytes.
B. Calculate the relative percent difference
(RPD) between first sample and replicate. The calculations must be documented,
and the target value must be close to the first value and have a RPD of less
than 20%.
C. Matrix-specific QC
samples indicate the effect of the sample matrix on the precision and accuracy
of the results generated using the selected method. The information from these
controls is sample/matrix specific and would not normally be used to determine
the validity of the entire batch.
D. For methods that include one (1) to twenty
(20) targets, spike all components.
E. For methods with more than twenty (20)
targets, randomly spike at least sixteen (16) components.
Sample Duplicate
A. Analyze the same sample twice, using two
separate preparations. The sample should be chosen at random and run together
on the same analytical run.
B.
Calculate the relative percent difference (RPD) between first sample and
replicate. Calculations must be documented, and the target value must be close
to the first value and have a RPD of less than twenty percent (20%).
Variability may be introduced during sample preparation. To
account for this, if more than one staff member is prepping samples, each staff
must also prepare and analyze a sample matrix duplicate for each set of
prepared samples.
