See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/226755351

Proficiency test exercise on the determination

of natural levels of radionuclides in mushroom
reference material
ARTICLE in ACCREDITATION AND QUALITY ASSURANCE MAY 2007
Impact Factor: 0.97 DOI: 10.1007/s00769-007-0260-y

CITATIONS

READS

49

5 AUTHORS, INCLUDING:
Shahida Waheed

Naila Siddique

18 PUBLICATIONS 177 CITATIONS

Pakistan Atomic Energy Commission

SEE PROFILE

66 PUBLICATIONS 592 CITATIONS

SEE PROFILE

M. Rossbach
Forschungszentrum Jlich
87 PUBLICATIONS 540 CITATIONS
SEE PROFILE

Available from: Naila Siddique

Retrieved on: 23 February 2016

Accred Qual Assur (2007) 12:311316

DOI 10.1007/s00769-007-0260-y

PRACTITIONERS REPORT

Proficiency test exercise on the determination of natural levels

of radionuclides in mushroom reference material
Shahida Waheed Asma Rahman Naila Siddique
Shujaat Ahmad Matthias Rossbach

Received: 30 May 2006 / Accepted: 20 January 2007 / Published online: 29 March 2007

Springer-Verlag 2007

Abstract The results from a proficiency test exercise for

the determination of natural levels of radionuclides in
mushroom reference material carried out by laboratories in
six different countries are discussed. These laboratories
submitted data on seven radionuclides, namely 134Cs,
137
Cs, 40K, 90Sr, 241Am, 226Ra, and 210Po. However, only
the data received for three of these radionuclides could be
statistically evaluated to obtain consensus mean values and
confidence intervals (at a significance level of 0.05). The
contents of 134Cs, 137Cs and 40K were quantified as 4.4,
2,899 and 1,136 Bq/kg, respectively.
Keywords Accuracy
Mushrooms
Precision

Proficiency test
Radionuclides

Reference materials (RMs)

Introduction
International trade is currently facing tough competition
due to strict legislation enforced by the World Trade
Organization (WTO), which stipulates certain levels of
quality for all processes and products for export. In the
light of these challenges, credible laboratories are required
to achieve and maintain quality standards which help

S. Waheed
A. Rahman
N. Siddique (&)

S. Ahmad
Chemistry Division, Directorate of Science,
Pakistan Institute of Nuclear Science and Technology
(PINSTECH), P.O. Nilore, Islamabad 45650, Pakistan
e-mail: nailasiddique@yahoo.ca; naila@pinstech.org.pk
M. Rossbach
Department of Nuclear Sciences and Applications,
IAEA, P.O. Box 100, Vienna, Austria

industry gain the confidence of the world market by calibrating performance through quality indicators. International bodies such as EURACHEM, IUPAC and ISO have
made considerable efforts to establish guidelines for a
general quality system that can be applied in analytical
laboratories to trace and document their results in such a
way that compatibility amongst laboratories can be obtained [1, 2]. A regular independent assessment of the
technical performance of a laboratory is recommended as
an important means of assuring the validity of analytical
measurements, not only as part of an overall quality
management strategy, but also to demonstrate competence
and to promote the confidence of national and international
customers. The use of reference materials (RMs) for reliable chemical analysis is an integral part of the work of
certified laboratories to maintain the accuracy and precision of their analytical results and to follow analytical
quality assurance and control procedures [3]. For this reason, there is a permanent requirement for participation in
proficiency tests (PTs) and interlaboratory comparison
exercises [4]. A diverse inventory of RMs is required, as
RMs are used in chemical analysis by certified laboratories,
and matrix RMs provide a more realistic approach to the
validation of the data being characterized. Such RMs exhibit a corresponding analytical archetype and interferences identical to the sample under investigation [5].
For more than three decades the International Atomic
Energy Agency has been producing natural matrix RMs.
These IAEA RMs are distributed for worldwide and regional intercomparison exercises amongst member states
through its Analytical Quality Control Services (AQCS).
These exercises are used to identify and harmonize analytical intricacies in laboratories by instigating corrective
actions, and are also useful for characterizing trends in
analytical performance over the years [6]. Moreover, these

123

312

exercises provide the IAEA member states with different

types of RMs and CRMs in sufficient quantities to serve
their needs [7]. The IAEA recently designed an interregional project, INT/1/054, entitled, Preparation of Reference Materials and Organization of Proficiency Test
Rounds. The aim of this project on the preparation,
homogenization and characterization of reference material
for PTs was to serve as a pilot study for participant
countries organizing such tests in their respective countries.
A PT was organized within the framework of this project.
This exercise also served to estimate the proficiencies of
the analytical laboratories of the countries participating in
this project at determining different constituents, such as
different radionuclides and heavy metals.
Following discussions, mushrooms were selected as a
suitable material for conducting PTs. Mushrooms have a
well-established place in European cuisine and are also
gaining popularity at the international level as an important
source of dietary protein. However, they accumulate not
only heavy and alkaline metals but also significant amounts
of 137Cs from the environment. Six countries viz. Brazil,
Hungary, Korea, Pakistan, Poland and Syria participated
and submitted data in this PT exercise on the determination
of radionuclides in edible mushroom material.
The neutron activation analysis (NAA) laboratory at the
Chemistry Division, Directorate of Science of the Pakistan
Institute of Nuclear Science and Technology (PINSTECH),
a premier research institute in the country, is one of the few
laboratories in Pakistan to obtain formal national accreditation by the Pakistan National Accreditation Council
(PNAC), under ISO/IEC-17025. To maintain confidence in
our analytical capabilities, the laboratory has been a regular
participant in IAEA Analytical Quality Control exercises
[811]. Our group at Chemistry Division was asked by the
IAEA to compile and publish the results from this PT. The
results from this exercise are presented here.

Materials and methods

Wild mushrooms, Xerocomus badius, family Boletaceae,
were selected for the study and were collected from
Poland, where they are easily available in the forests. More
than 200 kg of fresh mushrooms were collected. The
mushrooms were first cleaned and dust, soil, and attached
mosses were removed. The part of each mushroom that was
embedded in the soil was cut and removed. Mushrooms
were cut into small pieces and air-dried in a dryer at a
temperature of 2560 C according to the standard procedure used by food concentrate producers. Dried mushrooms were milled in a centrifugal mill and sieved.
Particles <1 mm diameter were collected. The mushroom
material was then sent to the Institute for Reference

123

Accred Qual Assur (2007) 12:311316

Materials and Measurements (IRMM) Laboratory, in Geel,

Belgium, for final preparation. Further processing of this
material was performed at this institute by the participants
in the project under the guidance of IAEA and IRMM
experts. The material was further milled using a ZrO2
planetary ball mill and sieved through 125 lm sieves and
homogenized by mixing. The homogenized samples were
then distributed in different bottles. The homogeneity of
the sample for 134Cs and 137Cs was checked in six replicate
samples of 5 g in 510 randomly selected bottles. INOVA
was used to check the homogeneity within the randomly
selected bottles (two subsamples from each bottle and six
replicates from each subsample). The results obtained
showed a difference of <5% between the activities measured for 134Cs and 137Cs, and samples of size 5 g were
therefore considered to be homogeneous. Characterization
of the samples was also carried out in IRMM, Belgium,
using sieve analysis, optical microscopy, and laser diffraction techniques, while water contents were determined
by Karl Fischer titration, water activity determination, and
sorption isotherm.
The participants were supplied with bottles each containing 20-g samples and were asked to analyze the samples in selected laboratories in their countries. A request
was made to the participants to perform at least three (but
preferably six) independent determinations and to report
the mean value, uncertainty and limit of detection for each
radionuclide. Each laboratory was assigned a code number,
thus ensuring anonymity. Since the moisture content can
vary with the ambient humidity and temperature of the
region where the analysis is performed, the participants
were requested to determine a separate subsample (not
used for analysis) by drying to a constant weight at 50 C
(usually at least 24 h) and report all the results on a dry
weight basis. The results from six different countries were
received. Three laboratories determined 134Cs, while all six
determined 137Cs and 40K. Only single laboratories reported 90Sr, 241Am, 226Ra and 210Po values, so the results
for these radionuclides are only included in this report for
reference purposes.

Evaluation of the results

Results for three radionuclides (i.e., 134Cs, 137Cs, and 40K)
in the mushroom reference material were reported by three
or more laboratories, and so they could be subjected to
statistical evaluation. The details of the individual laboratory mean values, reported uncertainties (absolute and
relative), sample masses and moisture contents are presented in Tables 1, 2, and 3 for 134Cs, 137Cs and 40K,
respectively. The sources of uncertainties included errors in
weighing, spectral peak counts, background counts and

Accred Qual Assur (2007) 12:311316

313

Table 1 Comparison of 134Cs results against the consensus values and acceptance tests (accuracy and precision) in mushroom reference material
Lab. Mean
Uncertainty Rel.
Mass Dry/
code (Bq/kg) (Bq/kg)
uncertainty (g)
wet
(%)
ratio
C1

3.70

Analyst/
Rel.
consensus bias
ratio
(%)

z-score u-score Accuracy criteria

Precision Status
p criteria
jxave  xi j 1:95 u2ave u2i (%)

0.90

24.32

NR

NR

0.84

15.5 0.8

0.6

0.7

2.2

30.8

Pass

C2

C3

Pass

C4

4.14

0.93

22.37

3.75

0.934 0.95

5.5

0.3

0.2

0.2

2.2

29.3

C5

5.30

2.10

39.62

9.63

0.930 1.21

21.0

1.1

0.4

0.9

3.97

43.9

Fail

C6

NR not reported

Table 2 Comparison of 137Cs results against the consensus values and acceptance tests (accuracy and precision) in mushroom reference material
Lab. Mean
Uncertainty Rel.
Mass
code (Bq/kg) (Bq/kg)
uncertainty (g)
(%)

Dry/
wet
ratio

Analyst/
Rel. z-score u-score Accuracy criteria
Precision Status
p criteria
consensus bias
2
2
jxave  xi j 1:95 uave ui (%)
ratio
(%)

C1

2,680.0 170.0

6.3

NR

NR

0.92

7.6 1.1

0.8

218.9

461.6

9.3

Pass

C2

2,710.0 140.0

5.2

NR

NR

0.93

6.5 1.0

0.8

188.9

429.1

8.6

Pass

C3

3,192.0 27.0

0.8

37.50 0.929 1.10

10.1 1.5

1.5

293.1

354.0

6.9

Pass

C4

3,039.5 30.3

1.0

3.75

0.934 1.05

4.8

0.7

0.7

140.6

354.8

6.9

Pass

C5

2,948.7 98.3

3.3

9.80

0.932 1.02

1.7

0.3

0.2

49.8

391.4

7.6

Pass

C6

2,823.2 36.5

1.3

2.20

NR

2.6 0.4

0.4

75.7

356.7

7.0

Pass

0.97

NR not reported
Table 3 Comparison of 40K results against the consensus values and acceptance tests (accuracy and precision) in mushroom reference material
Lab. Mean
Uncertainty Rel.
Mass Dry/
code (Bq/kg) (Bq/kg)
uncertainty (g)
wet
(%)
ratio

Analyst/ Rel.
consensus bias
ratio
(%)

z-score u-score Accuracy criteria

Precision Status
p criteria
2
jxave  xi j 1:95 u2ave ui (%)

C1

1,130.0 80.0

7.1

NR

NR

1.00

0.5

0.1

0.0

5.7

243.8

12.2

Pass

C2

1,001.0 50.0

5.0

NR

NR

0.88

11.9 1.2

1.1

134.7

217.4

11.1

Pass

C3

1,140.0 22.0

1.9

37.50 0.929 1.00

0.4

0.0

0.0

4.3

202.4

10.1

Pass

C4
C5

1,319.3 53.4
1,183.7 95.5

4.0
8.1

3.75
9.80

0.934 1.16
0.932 1.04

16.2
4.2

1.6
0.4

1.5
0.3

183.6
48.0

220.0
260.5

10.7
12.8

Pass
Pass

C6

1,040.0 136.0

13.1

NR

NR

8.4

0.8

0.5

95.7

311.7

16.4

Pass

0.92

NR not reported

uncertainties in concentrations of radionuclides used when

calculating activities. Results for a further four radionuclides, 90Sr, 241Am, 226Ra, and 210Po, presented in Table 4,
were reported by single laboratories only and hence no
consensus values of these radionuclides were available for
further evaluation.
The original data for 134Cs, 137Cs, and 40K received
from the participating laboratories were entered into the
computer program Histo Version 2.1 (IAEA) in the
required format. Four outlier tests (i.e., Dixon, Grubbs,
skewness and kurtosis) were applied to these data sets. All
values for these three radionuclides were accepted by the
software. Mean values or consensus values, absolute

standard deviations, relative standard deviations, standard

errors, medians and ranges of values for these three radionuclides in the IAEA mushroom reference material are
reported in Table 5. The results were further evaluated to
assess the difference between the consensus value and the
reported value using the following parameters:
1.

Relative bias (brel) between the analysts value and the

consensus mean is expressed as a percentage, given by
the expression [5]:

brel

xave  xi 100%
xave

123

314

Accred Qual Assur (2007) 12:311316

Table 4 Data for additional radionuclides reported by single laboratories in mushroom reference material
Radionuclide

241

210

226

90

Lab. code

C6

C1

C6

C5

Am

Po

Ra

Sr

No. of results averaged

Mass (g)

2.2

NR

2.2

9.63

Dry/wet ratio

NR

NR

NR

0.93

Mean (Bq/kg)

2.46

5.01

35.44

0.58

Uncertainty (Bq/kg)

0.26

1.11

1.3

0.11

Relative uncertainty (%)

10.43

22.16

3.67

18.97

LOD (Bq/kg)

1.68

NR

11.9

0.3

NR not reported

where xi is the arithmetic mean of the reported value of the

radionuclide concentration in the sample (laboratory value)
and xave is the consensus or mean value.
2. The z-score value giving each participant a normalized
performance score for bias is calculated according to
the equation [5]:
z

xi  xave
rb

j z j 62 = Satisfactory results
2\ j z j\3 = Results are questionable
j z j >3 = Unsatisfactory results.

*Confidence intervals are for a

significance level of 0.05

123

xi  xave
En or u p
u2i u2ave
where xi and xave are defined above, ui is the uncertainty in
the participants result and uave is the uncertainty in the
consensus value. The u-scores were compared with the
critical values listed in the t-statistic tables to determine
whether the reported result differs significantly from the
expected value at a given level of probability [13]. It
should be noted that the choice of the significance level is
subjective. For this PT, we have set the limiting value for
the u test parameter to 1.95 to determine if a result passes
the test (u < 1.95), i.e.,
j En jorj u j 6 1.95=Satisfactory results
j En jorj u j > 1.95=Unsatisfactory results.
4.

where xi and xave are as defined above, and rb is the target

value for the standard deviation (r). These have been
assigned on the basis of the reproducibility standard
deviation (the standard deviation of the consensus mean
after outlier rejection, which expresses the interlaboratory
precision). For this exercise, the calculated values are 18.9,
6.9, 9.9% for 134Cs, 137Cs and 40K, respectively. The
acceptance criterion for the z-scores is:

Table 5 Summarized results

for the radionuclides 134Cs,
137
Cs and 40K with confidence
intervals in mushroom reference
material

This method of assessing laboratories has been accepted as

standard for ISO/IEC Guide 43-1 [5, 12].
3. The u test score or the En number or u-score is given
by [5]:

A result passes the accuracy criteria at 95% confidence

level if [5]:

jxave  xi j6 1:95

q
u2ave u2i

where xi, xave, ui, and uave are defined above.

5. For the precision criterion in this study, the result
passes if [5]:

s

 2
uave 2
ui
100% 6 2rrep

xave
xi
where xi, xave, ui, and uave are defined above, and rrep is the
reproducibility standard deviation as given in the table for

Radionuclide

134

137

4

No. of laboratories

No. of results averaged

22

22

Range of values
(Bq/kg)

3.705.30

2,680.03,192.0

1,001.01,319.3

Overall mean (Bq/kg)

4.38

2,898.9

1,135.7

Abs. SD (Bq/kg)

0.83

198.7

112.6

Relative SD (%)
Standard error (abs)

18.9
0.50

6.9
81.1

9.9
46

Standard error (%)

10.9

2.8

4.0

Confidence interval
(Bq/kg)*

3.45.3

2,7403,058

1,0461,226

Median (Bq/kg)

4.14

2,885.9

1,135.0

Cs

Cs

Analysts were requested to report their results together

with the corresponding combined standard uncertainties.
The evaluation procedure used to calculate critical values
for accuracy and precision involved the reported value and
its corresponding combined standard uncertainty. To
determine the acceptance range for a result to pass the
accuracy criterion for this PT, we have set the two-tailed
value for the Students t distribution to the 0.05 significance level. The precision criterion defines the maximum
acceptable uncertainty that could be assigned to the reported value and this was set to twice the reproducibility
standard deviation (which expresses the interlaboratory
precision) for this exercise, again using 95% confidence
statistics. These fairly large limits were applied while
bearing in mind the small number of participating laboratories, reporting a small number of values.
As the material used was being analyzed for the first
time, no certified or reference property values were available and the target values were derived from the consensus
means, which are the mean values obtained from this PT
exercise after outlier rejection. Hence only those three radionuclides which were reported by more than one laboratory (i.e., 134Cs, 1374Cs and 40K) could be assigned
property values. Figures 1, 2 and 3 present the distribution
of results expressed as the analyst/consensus value ratios
for 134Cs, 137Cs, and 40K, respectively, in the mushroom
reference material. The results are organized according to
the laboratory code in ascending order. The error bars used
in the figure represent the laboratory ratio combined
standard uncertainty for the ratio. The horizontal dashed
lines represent two standard deviations of the overall
mean of these ratios. Figure 4 represents the information
values as reported by single laboratories for the four radionuclides 241Am, 210Po, 226Ra, and 90Sr.
Only three laboratories reported analytical results for
134
Cs. All seven values provided were accepted on the
basis of outlier tests, resulting in a relatively large uncertainty. Because of this large uncertainty all laboratory
mean results passed the z-score, u test score and accuracy
criteria. However the individual large uncertainty of laboratory C5 (39.62%) is reflected in the failed precision test
(Tables 1, 5).
All six participating laboratories provided results for
137
Cs and so 22 values were available for this radionuclide.
The concentration of this radionuclide was the highest and
the overall standard deviation of laboratory means was the
lowest amongst the three selected radionuclides, reflecting
the fact that the analytical results obtained for this radionu-

2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
C1

C2

C3

C4

C5

C6

Laboratory Code

Fig. 1 Results for 134Cs expressed as the analyst mean/consensus

mean ratio in IAEA mushroom reference material

Analyst Results/ Consensus Mean

Results and conclusions

2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
C1

C2

C3

C4

C5

C6

Laboratory Code

Fig. 2 Results for 137Cs expressed as the analyst mean/consensus

mean ratio in IAEA mushroom reference material

Analyst Results/ Consensus Mean

z-scores using 95% confidence statistics. If a particular result fails to satisfy any or all of the above tests, it is rejected.

315

Analyst Results/ Consensus Mean

Accred Qual Assur (2007) 12:311316

2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
C1

C2

C3

C4

C5

C6

Laboratory Code

Fig. 3 Results for 40K expressed as the analyst mean/consensus mean

ratio in IAEA mushroom reference material

clide were the best. Using 95% confidence criteria, all results
passed in terms of accuracy and precision (Tables 2, 5).
A similar number of results were received for 40K from
all six participating laboratories. The uncertainties were
slightly higher than for 137Cs, as was the range of values.
However all values reported were acceptable on the basis
of outlier tests. Results from accuracy and precision tests

123

316

Accred Qual Assur (2007) 12:311316

100.0

Radionuclide Activity

Ra-226

10.0
Po-210
Am-241
1.0
Sr-90

0.1

Radionuclide

Fig. 4 Results for 241Am, 210Po, 226Ra, and 90Sr in IAEA mushroom
reference material

were acceptable for all laboratory means using 95% confidence statistics (Tables 3, 5).
The final consensus means and associated confidence
intervals (at a significance level of 0.05) for 134Cs, 1374Cs and
40
K are presented in Table 5. Recommended values could
not be formulated because of the very limited data available.
Acknowledgments The authors are thankful to Dr. Zbigniew,
RADECKI Technical Coordinator of the IAEA project INT/1/054
entitled Preparation of Reference Materials and Organization of
Proficiency Test Rounds, for providing the opportunity to compile
the international data. We also acknowledge Ms. Halina PolkowskaMotrenko, from the Institute of Nuclear Chemistry and Technology,
Warsaw, Poland, for the collection of this material. Finally, all of the
laboratories who participated in this exercise are also acknowledged.

References
1. Thompson M, Wood R (1995) IUPAC Technical Report: Harmonized guidelines for internal quality control in analytical
chemistry laboratories. International Union of Pure and Applied
Chemistry, Zurich

123

2. Bode P (1997) From misconception to a must: the measured

merits of total quality management and accreditation in INAA.
J Radioanal Nucl Chem 215:5157
3. Okamoto K, Yoshinage J (1999) Use of matrix reference materials. Wiley, Berlin
4. ISO/IEC (1997) ISO Guide 43-1: Proficiency testing by interlaboratory comparisons, Part 1. International Standard Organization, Geneva
5. Bleise AR, Smodis B, Glavic-Cindro D, Parr RM (2001) The
updated IAEA database of natural matrix reference materials.
J Radioanal Nucl Chem 248:205209
6. Coquery M, Carvalho FP, Azemard S, Horvat M (1999) The
IAEA worldwide intercomparison exercises (19901997): determination of trace elements in marine sediments and biological
samples. Sci Total Environ 237/238:501508
7. Povinec PP (1996) IAEA-MEL Report 5/96: Future challenges of
IAEAs AQCS programme. IAEA-MEL, Monaco
8. Mannan A, Waheed S, Rahman A, Ahmad S, Qureshi IH (1993)
Multi-element analysis of coastal marine sediment, on IAEA
proposed reference material (IAEA 356). Geostand Newsl
17(1):223226
9. Waheed S, Waseem M, Rahman A, Ahmad S (2001) Instrumental
neutron activation analysis of estuarine sediment as a proposed
reference material for environmental studies. Geostand Newsl
25:137144
10. Waseem M, Rahman A, Zaidi JH, Waheed S, Ahmad S (2001)
Neutron activation analysis of IAEA proposed certified reference
material IAEA-407 (North sea fish homogenate). J Radioanal
Nucl Chem 248:205209
11. Waheed S, Rahman A, Daud M, Zaidi JH, Ahmad S (2004)
INAA methodology using low power reactor for the characterization of IAEA-433; Coastal marine sediment as a potential
reference material. Radiochim Acta 92:939944
12. Thompson M, Wood R (1993) The international harmonized
protocols for the proficiency testing of (chemical) analytical
laboratory studies, IUPAC/ISO/AOAC. J Pure Appl Chem
65:21232144
13. Brookes CJ, Betteley IG, Loxton SM (1979) Fundamentals of
mathematics and statistics. Wiley, New York