|
APG provides an array of samples. The majority of these samples fall into three categories.
1. Calibration samples
2. Quality control standards
3. Proficiency testing samples
Proper use of this information provides a solid foundation to verify the quality of work in your daily operations.
|
 |
|
First Level: Proper Calibration Curve
A proper calibration curve is essential to generating reliable data. Without a
proper calibration curve, little useful information can be extrapolated from
data. A proper calibration curve should provide consistent results across
its range. Regardless of its origins, a proper calibration curve is essential
to defending the quality of laboratory results. The details of a calibration
curve can be found in a previous article written
in the December 2003 eNewsletter .
Second Level: Independently Prepared Quality Control Standards
This is a known standard and thus information regarding it is generally provided to
the person performing the analysis. An APG QC Standard includes a Certificate of
Analysis (C of A), which provides the information needed to ensure confidence in
your analysis. Among this information is a column labeled Prepared Value, which
is the gravimetric true value of the given analyte. Also contained in a C of A is a
Certified Value column which is the value obtained by APG from our own internal testing.
One of the last pieces of information on a certificate of analysis is a column
labeled Acceptance Range. In order to truly verify a calibration, the quality control standard
should not be created from the same stock because the same issues may be present in both. Once
run against a calibration, the independent quality control sample verifies that
the calibration is correct.
If the value of the known sample falls within the acceptance limits you can be reasonably
certain in the accuracy of your calibration curve. Results near the edges of the acceptance
range may be a warning of potential issues and thus should be investigated. Analysts will
often run two QC samples, one directly after the calibration and one after all other
samples. The second QC verifies that there has been no drift in the instrument during the run.
In addition to running known samples to verify a calibration, QCs can also be useful for training
of analysts in new methods. They provide a quick means to test a new analyst's ability to
perform a method.
Third Level: Proficiency Testing Samples
A PT sample shows your labs ability to test a sample which is unknown. Once
your lab runs the samples and produces results, the data is sent back to the proficiency
testing providers for evaluation. Proficiency testing reports contain valuable information
that is best obtained from participating in a study with many laboratories (i.e. round
robin testing). The inter-laboratory percent recovery compares recoveries between all
participating laboratories for each parameter.
Z Scores are another useful tool. Z Scores measure the relative distance in standard deviations
from the mean of given data set. A Z Score less than one normally indicates a method that is
under control and running properly. A Z Score of two to three is passing for a wastewater
study but shows signs of concern and should be reviewed. A Z Score of above three is
unacceptable for wastewater and a Z Score above two is unacceptable for drinking water. Z Scores
can also be used over time to monitor a method. If your lab notices its Z Score drift
over time, noticeably in one direction, it could indicate a method that is not under
control and needs to be investigated.
Putting It All Together
Calibration standards, quality control
standards and proficiency testing samples are at the very core of a labs overall quality control policy.
Used in combination they ensure results which
are defensible and repeatable. Running a proper calibration, QC samples and participating
in a PT program builds confidence that results are defensible.
|
