Transcript PowerPoint-Präsentation

Dr. K. Luden
Lab Quality Assurance by Interlaboratory
Comparisons
Proficiency Testing
Dr. Katrin Luden
Warschau Nov. 2008
1
Dr. K. Luden
Drinking water directive 98/83/EC
Annex III
Specifications for the analysis of Parameters
Each Member State must ensure that any laboratory at
which samples are analysed has a system of analytical
quality control that is subject from time to time to
checking by a person who is not under the control of the
laboratory and who is approved by the competent
authority for that purpose.
Warschau Nov. 2008
2
Dr. K. Luden
German drinking water directive
§ 15 Methods of Analysis and Laboratories
(4) The nessessary analysis including sampling are to be
conducted by laboratories that work by “generally accepted
rules of the technology” (allgemein anerkannte Regeln der
Technik), have a system of internal quality control,
participate at least once a year in external quality
assurance schemes, have sufficiently qualified personell
and are accredited by a generally recognized accreditation
body. The responsible highest federal authority will publish
a list of laboratories in its jurisdiction.
(5) The authorities (of the federal state) independent of the
labs will have to check regularly wether the conditions of
(4) are still fullfilled.
Warschau Nov. 2008
3
Dr. K. Luden
Proficiency testing
Participation in proficiency testing programs allows
laboratories to:
 identify areas where improvement in their testing and
measurement methods is needed
 identify further training of their staff
 foster confidence in the performance of their testing and
measurements
 assure laboratory performance in their accredited test and
calibration results
Warschau Nov. 2008
4
Dr. K. Luden
Jungle of standards for PT / EQA
There are several standards and guidelines for providing PT / EQA schemes
(in chronological order)
• ISO/IEC Guide 43-1:1996 Proficiency testing by Interlaboratory Comparisons –
Part 1: Development and Operation of Proficiency testing Schemes
• ISO 13528:2005 Statistical Methods for use in Proficiency Testing by
Interlaboratory comparisons
• IUPAC Technical Report 1/2006 The International Harmonized Protocol for the
Proficiency Testing of Analytical Chemistry Laboratories
• ILAC-G13:2007 ILAC Guidelines for the Competence of Providers of Proficiency
Testing Schemes
• ISO/TS 20612 Water Quality – Interlaboratory comparisons for proficiency testing
of analytical chemistry laboratories
• ISO/IEC CD 17043:2008 Conformity Assessment – General Requirements for
Proficiency Testing
Warschau Nov. 2008
5
Dr. K. Luden
External Quality Assurance
Recommendation of Federal Environmental Agency
2002 - Microbiology
Warschau Nov. 2008
6
Dr. K. Luden
External Quality Assurance
Recommendation of Federal Environmental Agency
2002 - Microbiology
successfull participation in
proficiency testing schemes
Parameter
E. coli/Coliform bacteria
Enterococci
Colony Count
2x / year
Clostridium perfringens
Legionella
Pseudomonas aeruginosa
1x / year
Enteropathogenic Viruses
Cryptosporidium parvum
International PTS
Warschau Nov. 2008
7
Dr. K. Luden
External Quality Assurance
Recommendation of Federal Environmental Agency
2003 - Chemistry
successfull participation in
proficiency testing schemes
Parameter
All parameters offered by a
laboratory
1x within 2-3 years
 Proficiency testing (PT) according to DIN 38402-45
Warschau Nov. 2008
8
Dr. K. Luden
German drinking water PT providers
 Microbiology and
“hygenic-chemical PT
 Chemistry PT
 Chemistry PT
 Chemistry PT
Regular meetings, harmonized schedules, identical
statistical evaluation → comparability for labs and authorities
Warschau Nov. 2008
9
Dr. K. Luden
DIN 38402-45
 Standard may be applied if:
 Parameter analysed can be considered continuously
measureable
true for most chemical and phisico-chemical parameters
not allways true for biological and/or microbiological
parameters
 Application in Chemistry: Yes
 Application in microbiology: by way of trial / mostly
Warschau Nov. 2008
10
Dr. K. Luden
Microbiology PT
The governmental Institute of Public Health of Lower
Saxony (NLGA) is Proficency Test (PT) provider for
microbiological measurements in drinking water in
Germany
Every year:
 4 PTS for E. coli/coliform bacteria, Enterococci, and
Colony Counts
 2 PTS for Pseudomonas aeruginosa, Clostridium
perfringens and Legionella
 1 PTS for EU bathing water (E. coli, coliforme B.)

1 PTS for selected “hygenic-chemical” parameters
Warschau Nov. 2008
11
Dr. K. Luden
Microbiology PT
Parameter
E. coli-coliform
bacteria
Colony Count
Enterococci
Clostridium
perfringens
Pseudomonas
aeruginosa
Qualified methods
1986 1999
presence/absence test
+
DIN EN ISO 9308-1
(+)
Colilert®-18/Quanti-Tray®
German DWD 1990
+
DIN EN ISO 6222
EN ISO 7899-2
Chromocult®-Enterococci-Agar
DWD (mCP-Agar)
TSC
2000 2001 2002 2003 2004 2006 2007 2008
+
(+)
+
(+)
+
(+)
+
+
(+)
+
+
+
+
+
+
+
+
(+)
+
+
+
+
+
+
+
+
(+)
+
+
+
+
+
+
+
+
+
+
+
(+)
(+)
(+)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
(+)
DIN EN ISO 9803-3
(+)
+
+
DIN EN ISO 7899-2
DIN EN ISO 7899-1
(+)
(+)
+
+
+
+
(+)
EN 12780/ISO 16266
+
+
+
(+)
+
+
+
+
+
(+)
+
+
+
+
+
+
+
+
+
+
Legionella spec. GVPC (ISO 11731-2)
EC bathing water directive
Feacal coliform
bacteria
E. coli-coliform
bacteria
Intestinal
Enterococci
Warschau Nov. 2008
+
12
+
Dr. K. Luden
Microbiology PT
Parameter
E. coli-coliform
bacteria
Colony Count
Enterococci
Clostridium
perfringens
Pseudomonas
aeruginosa
Qualified methods
1986 1999
presence/absence test
+
DIN EN ISO 9308-1
(+)
Colilert®-18/Quanti-Tray®
German DWD 1990
+
DIN EN ISO 6222
EN ISO 7899-2
Chromocult®-Enterococci-Agar
DWD (mCP-Agar)
TSC
2000 2001 2002 2003 2004 2006 2007 2008
+
(+)
+
(+)
+
(+)
+
+
(+)
+
+
+
+
+
+
+
+
(+)
+
+
+
+
+
+
+
+
(+)
+
+
+
+
+
+
+
+
+
+
+
(+)
(+)
(+)
+
+
+
+
+
+
+
+
+
+
+
+
+
+
(+)
DIN EN ISO 9803-3
(+)
+
+
DIN EN ISO 7899-2
DIN EN ISO 7899-1
(+)
(+)
+
+
+
+
(+)
EN 12780/ISO 16266
+
+
+
(+)
+
+
+
+
+
(+)
+
+
+
+
+
+
+
+
+
+
Legionella spec. GVPC (ISO 11731-2)
EC bathing water directive
Feacal coliform
bacteria
E. coli-coliform
bacteria
Intestinal
Enterococci
Warschau Nov. 2008
+
13
+
Dr. K. Luden
PT samples - two options
Liquid samples
Freeze dried material
Bacterial solution in defined
mineral medium
Bacteria adsorbed to milk
powder
Very realistic samples
Handling exactly like a real
sample
A solution has to be prepared
and first step of handling
procedure is different from
real samples
Stability 4-6 days after
shipment at <10°C
Stability at -20°C weeks or
months
Germany
United Kingdom, France,
Sweden, Australia...
Warschau Nov. 2008
14
Dr. K. Luden
Preparation at NLGA
Warschau Nov. 2008
15
Dr. K. Luden
Shipping
300 - 500
participants/PT
Registered:
> 700 Labs from
Germany
Austria (26)
Switzerland (8)
Lithuania (2)
Hungary (2)
Great Britain (1)
Luxembourg (1)
Warschau Nov. 2008
16
Dr. K. Luden
Schedule of a proficiency testing scheme
 Registration of participants
 Confirmation of the registration (4 weeks before PT)
 Preparation of the samples and shipment
 Internal quality control (two weeks before and 6 days
after shipment)
 Reporting of the results (within 16 days of shipment)
 Evaluation of the results usually within 4 weeks after
the deadline for reporting
Warschau Nov. 2008
17
Dr. K. Luden
What do the labs get?
 Report on the PT design, special problems, interesting
results etc.
 Certificate and evaluation report
(target value as consensus value of participants results)
 Quality control charts about the influence of temperature
and the duration of shipping
 Z(u)-Scores of all measurements in tables
 Z(u)-Scores in graphic form
Warschau Nov. 2008
18
Dr. K. Luden
Statistics of DIN 38402-45
 Robust statistics (no outlier tests); consensus mean
(used as assigned value) calculated by Hampel
Estimator
m-4,5s
m-3s
m-1,5s
m
m+1,5s
m+3s
m+4,5s
 ZU-scores “normalize” the results for comparing
different levels of bacterial concentration
lZUl  2 - sufficient
lZUl  3 - questionable
lZUl > 3 - insufficient
( x  mt )
Z
st
Z-score:
Warschau Nov. 2008
19
Dr. K. Luden
Certificate
Name/Adress of PT-provider
Address of participant
Identification of PT (e.g. 3-2008)
Table of parameters,
successfully analyzed param.
And method used
Signatures of PT provider
Warschau Nov. 2008
20
Dr. K. Luden
Microbiology PT
ZU-score
Group B III-2008 E. coli (assigned value 21 CFU/100 ml)
Labcode
Warschau Nov. 2008
21
Dr. K. Luden
Interlaboratory comparisons
 method validation studies:
aptitude/qualification of the lab is prerequisite
reference material is fit for purpose
 to certify reference material:
method is supposed to be fit for purpose/valid
labs are expert labs
 Quality control/proficiency testing:
performance of a lab is tested, reference material and
method are supposed to be valid/fit for purpose
Warschau Nov. 2008
22
Dr. K. Luden
E. coli
PT 1-2003
45
40
E. coli CFU / 100 ml
35
30
25
20
15
10
5
X+3s
36
X+2s
386
X
60
23
X-2s
391
Warschau Nov. 2008
X-3s
522
E. coli / 100 ml
545
Labcode
149
279
116
460
367
284
469
442
328
219
459
437
340
56
178
543
445
406
162
12
35
183
3
428
5
R
R
R
R
R
R
R
R
R
R
0
Dr. K. Luden
Clostridium perfringens
CFU/100 ml
PT 2-2004
Labcode
Warschau Nov. 2008
24
Dr. K. Luden
Clostridium perfringens
Medium/method used
Recovery
rate
Columbia blood agar / membrane filtration
100 %
Columbia blood agar / membrane filtration no shaking of the sample
39 %
DEV-Nutrient agar / pour plate method
78 %
mCP Agar / Filtration (24 hours TNT)
43 %
mCP Agar / Filtration (24 hours 4°C)
43 %
mCP Agar / Filtration (PT participants mean)
11 %
Warschau Nov. 2008
25
Dr. K. Luden
Clostridium perfringens
Medium/method used
Recovery
rate
Columbia blood agar / membrane filtration
100 %
Columbia blood agar / membrane filtration no shaking of the sample
39 %
DEV-Nutrient agar / pour plate method
78 %
mCP Agar / Filtration (24 hours TNT)
43 %
mCP Agar / Filtration (24 hours 4°C)
43 %
mCP Agar / Filtration (PT participants mean)
11 %
Warschau Nov. 2008
26
Dr. K. Luden

Warschau Nov. 2008
27
Dr. K. Luden
Pseudomonas aeruginosa
In PT 1-2007
30 % of labs failed because
they reported a numerical
result for P. aeruginosa
when the strain provided
was not a P. aeruginosa
Possible explanation:
problems with confirmation
reaction - Ammonia
formation from acetamide
Warschau Nov. 2008
28
Dr. K. Luden
E. coli
A pure culture of E. coli was
used in PT 4-2004
Colilert (n = 80)
False results:
ISO 9308-1: 2,4 %
Colilert®-18: 18,8 %
ISO 9308-1 (n = 246)
Coliform
E. coli
delta
bacteria
68
7
61
0
25
25
30
54
24
15
16
1
21
22
1
24
25
1
Warschau Nov. 2008
29
E. coli
Coliform
bacteria
delta
0
66
62
74
0
0
74
30
48
48
50
53
19
24
27
66
70
70
78
78
83
83
43
50
53
53
59
21
25
45
66
4
8
4
78
83
9
13
2
5
3
6
2
1
18
Dr. K. Luden
Explanation?
ß-Galactosidase
ß-Glucuronidase
Warschau Nov. 2008
30
Dr. K. Luden
Conclusions
 Interaction between filter and nutrient agar might be
unfavorable for growth of the target organism
 Performance of the material combinations used has to be
checked (even if supplied with a certificate)
 Negative controls are as important as positive controls
 One can learn about the methods while testing
performance of the lab
 ....
Warschau Nov. 2008
31
Dr. K. Luden
Thank you for your attention!
Warschau Nov. 2008
32
Dr. K. Luden
Most likely errors - E. coli / coliforms
 Formal mistakes in the results: E. coli is part of the group
of coliforms, therefore the number of E. coli cannot be
larger than the number of total coliforms
 Commercial identification systems (that are not to be used
for the PT) might give conflicting results with the PT
method
 ISO 9803-1: 44°C Indole formation, wrong or not exactly
controlled incubation temperature (use a waterbath)
 Colilert®-18: measurements carried out at the wrong
wavelength; cavities not properly filled.
Warschau Nov. 2008
33
Dr. K. Luden
Most likely errors colony counts
 Contamination of the molten agar or the sample
 Temperature of the agar too high: a slight increase in
temperature might cause considerable decrease in
numbers of bacteria that are able to form colonies; too low
temperatures might lead to uneven distribution of bacteria
as the agar starts clumping and proper mixing is
prevented
 Volume of agar is too small
Warschau Nov. 2008
34