Article written by
Tom Coyner, President, Analytical Products Group
Last month, I started a series of articles on international vs U.S. PT programs. This month I want to talk
about multiple levels and frequency of testing. The Canadian PT program requires eight PT results a
year for each analyte. These must be completed as multiple levels so it might be two studies with four levels each
or four studies of two levels each. While in Germany we visited a drinking water PT program administrator
who does four studies a year at six levels each. Not exactly your normal NELAC requirement of two
studies of only a single level.
So who’s right? It all depends on why you are doing PT programs in the first place. If, as in the U.S., you
are using PT to judge failures and remove accreditation, it would be tough to sell a multi-level quarterly
PT program. The labs simply have too much at risk. If on the other hand you want to improve performance
and get useful data about how the labs are doing across a broad range of concentrations, then a multi-level program
is ideal. There is no question that the simple addition of a second level would allow you to analyze the
data as a Youden pair and provide volumes of information not available in a single level program. Youden was
a NIST statistician that developed the multi-level interlaboratory statistics for AOAC. Interestingly, the
USEPA WS and WP programs were both operated as two level, Youden pair programs for many years. These only
became single level programs as a result of EPA budget cuts not technical reasons.
If proficiency testing is done for laboratory quality improvement, multiple levels are really the way to go.
Given a sound set of interlaboratory results, a laboratory with poor performance can almost pinpoint the source
of the problem. The key to success is in the reporting structure of the PT program report. With a complete
interlaboratory analysis of the data from a study, patterns of performance and quality issues are highlighted.
APG is currently the only U.S. PT provider to offer a two level true interlaboratory program.
In January of 2006, APG will introduce a completely new system of PT testing where laboratories may run as many levels of
any sample they want in any study they want. Under this system APG customers who routinely run two levels will
simply order the second level as an additional level and these are available throughout the year. Other laboratories, such
as the Canadian labs, need to order the first level and as many additional levels that
they require. The number of levels and the number of studies can be selected by the customer. As you might suspect,
additional levels are less expensive than the first level. This is the most flexible customer friendly PT system
in the world.
|
 |
|
|
