|
Article written by Jessica Raney, Quality Control Group Leader, Analytical Products Group
Complete Article from Edition 8 APG eNewsletter
It is obvious that Proficiency Testing (PT) samples provide a judgment of competency - pass / fail.
By using round robin PT samples, you can compare yourself to other laboratories
using inter-laboratory statistics. In addition, they serve an often overlooked purpose.
PT Samples are an important part of an effective quality control program.
Analytical Quality Control
There are three main types of standards in any analytical quality control system. Each
is defined by its intended use and the information that must be
known about it.
-
Calibration
standards are used to establish the relationship between instrument
response and analyte concentration, therefore the true
value and identity must be known.
-
A quality control standard is a known standard
used to validate the calibration. The quality
control standard's true value and identity must be known as well as
some sort of measure of acceptable results. The measure is normally
provided as acceptance limits based on data generated during a round
robin testing study.
-
The PT sample is a blind sample used to evaluate
the quality control and quality assurance systems. The PT standard must
have a true value in order to evaluate the effectiveness of the quality
control and quality assurance systems.
Quality assurance includes the steps used to measure the effectiveness of
a quality control system. PT samples are one example of quality
assurance. Understanding the specific uses of calibration standards,
quality control standards, and proficiency testing samples is essential to
establishing and maintaining control of the analytical system. For a
further explaination of quality control standards, please refer to
the article
The Importance of Independently Prepared Quality Control Materials
.
PT Samples - A Judgment of Competency
One of the most widely recognized uses of PT samples is as
a judgment of laboratory competence. The US EPA and NELAC testing
programs are some familiar examples. In these programs, the true value
of the standard and historical data about the parameter of interest
determine the acceptance criteria. Regression constants derived
from historical data are applied to the true value of each parameter and
a mean and standard deviation are calculated. A confidence interval
is also generated around the calculated mean.
PT Samples - A Part of Peer Review
While the use of historical data as a judgment criteria accounts for method
bias and historical performance, it does not really allow for comparability
to other laboratories. Comparability is one of the important indicators of
a good quality assurance program. The inter-laboratory data
generated during a round robin study is essential to establishing comparability. The
PT sample is a peer review using the inter-laboratory data evaluations.
The inter-laboratory mean and standard
deviation are generated from the results reported by all of the laboratories
that tested the PT sample. Acceptance limits are calculated using a confidence interval
based around the mean.
The inter-laboratory study data will also provide two other
useful values.
-
The inter-laboratory average percent recovery is useful in evaluating your performance
on an individual parameter. This can help judge the
laboratory's accuracy.
-
The Z score is an even more useful tool generated during the study.
It allows for ranking of the participating laboratories
based on their reported result's proximity to the study mean.
In a wastewater (WP) study, a Z score of less than one indicates that
the laboratory is favorably comparable in analytical ability to its peers.
Z scores of two to three are cause for concern and further
evaluation, and a Z score greater than three is unacceptable. In a
drinking water (WS) study, a Z score of less than one is favorable,
one to two is average, and greater than two is unacceptable.
The Z score is essential for
evaluating the comparability of the laboratory to other labs. This is a
powerful marketing tool for highlighting a laboratory's analytical
ability versus competitors or other labs within
their organization.
PT Samples - A Quality Improvement Tool
A third function of the PT sample is as a quality improvement tool. Improvement
in Z scores indicates a comparability improvement. Because a PT sample
is unknown ("blind"), it is very useful in determining the laboratory's
ability to detect parameters and to find false positives.
PT samples also measure the quality control sample's performance. The PT
sample evaluates the entire analytical system and can find issues with quality
control before the quality control sample fails the acceptance criteria. The
PT sample is an independent third party check that evaluates
the analytical system.
Regular participation in PT studies is important to maintaining control of
the analytical method by allowing the laboratory to evaluate the
performance of the quality control standards. As part of an overall
quality control program, PT samples can be used to
train personnel, find issues in the analytical system, and prevent
problems before the affect data quality.
|
