www.
cenma
.cl
I. Lineamientos y tendencias
48
Materiales de referencia y comparaciones interlaboratorios
Points are then assigned for each result using the following
criteria;
The PT score for the parameter is then calculated from the four
samples as;
PT
Score
=
Total Points
4
×
20
All data used in this study were obtained fromCAEAL’s PT data-
base. The asbestos and the toxicity test groups were excluded
due to the different scoring procedures used for these test
groups. The following criteria were used to select data for use;
The first participation for a laboratory parameter combination
had to start after October 1994.This ensured that the first par-
ticipation event was being captured.
The laboratory parameter participation had to be continuous
for at least ten consecutive PT rounds following the first event
(five years).This restricted the data set to 84 parameters includ-
ing inorganics and organics in water, inorganics and organics
in soil, inorganics on filters, and microbiological. In total, 29,480
individual parameter scores were used (2,948 laboratory pa-
rameter combinations).
The purpose of this study is to determine if PT participation
improves laboratory performance. Thus, regardless of when a
laboratory started participation in a parameter, the first score
was assigned to study number one. Therefore, study one for
any record could be between January 1995 and January 1999.
III. Results
Most test groups displayed an obvious improvement over time
(e.g., figure 1). Others, however, displayed little change in per-
formance (figure 2).
When all parameters are combined, including the test groups
consisting of 1 to 5 parameters, there is an obvious trend to-
wards improved performance over the first few PT rounds.This
is coincident with a decreased variation in scores (figure 3).
IV. Discussion
This study demonstrates that, on average, the performance of
analytical laboratories, as measured by the CAEAL PT scoring
scheme, will improve over the first few participation events
(five to six), followed by a plateau. Although this is clearly dem-
onstrated in the combined data, the trend is not as clear on
a parameter level basis. Most individual parameters display
an improved performance; however, some parameters dem-
onstrate no observable change. The variable response for
individual tests is probably related to how well established a
procedure is, how rugged the methodology is and whether or
not a specific method is widely used (or regulated).There is in-
sufficient information in the CAEAL database to evaluate these
possibilities further.
A measured improvement in performance over the first few PT
rounds is not unexpected.Due to the cost of PT,the decision to
participate in proficiency testing is usually made as part of an
overall quality assurance effort, often aimed at the accredita-
Figure 3. Comparison of average PT score for all
parameters combined.
Figure 2. PT test groups where there was no
observable change in performance.
Figure 1. PT test groups where an observable
improvement in performance was evident.
1 2 3 4 5 6 7 8 9 10
82
84
86
88
90
92
94
96
PAHs inwater
PAHs insoil
OCPesticides
Average Score
Study
1 2 3 4 5 6 7 8 9 10
82
84
86
88
90
92
94
96
Metals inwater
Metals insoil
Major Ions
0 1 2 3 4 5 6 7 8 9 10 11
82
84
86
88
90
92
94
96
Inorganics inair
Microbiology
PCBs inoil
VOCs inwater
Average Score
Study
0 1 2 3 4 5 6 7 8 9 10 11
84
86
88
90
92
94
Average Score
Study
Ec/2
