Quality Assurance (QA), Quality Control (QC), Good

Lab Practices (GLP) in soil laboratories.

Christian Hartmann

To cite this version:

Christian Hartmann. Quality Assurance (QA), Quality Control (QC), Good Lab Practices (GLP) in
soil laboratories.. Doctoral. Japan. 2022. �ird-03854658�

HAL Id: ird-03854658

https://ird.hal.science/ird-03854658
Submitted on 16 Nov 2022

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
14 Nov. 2022 (10:30 – 11:00)
JICA training

Quality Assurance (QA)

Quality Control (QC)
Good Lab Prac6ces (GLP)
in soil laboratories.

Dr Chris(an Hartmann
IRD-France
GLOSOLAN steering commi2e
I – INTRODUCTION:

2
I – INTRODUCTION:

3
I – INTRODUCTION:

DEFINITIONS & CONCEPTS

QUALITY (Q):
Q.Assurance, Q. Control, Q. management….

4
I – INTRODUCTION:

DEFINITIONS & CONCEPTS

… .
IED
IF Control, Q. management….
QUALITY (Q), Q.Assurance,LQ.
P
SIM

5
QUALITY: of a product…
A product has good quality when:
it fits with the requirements specified by the client.

6
QUALITY: of a product…
A product has good quality when:
it fits with the requirements specified by the client.

Client:

7
QUALITY: of a product…
A product has good quality when:
it fits with the requirements specified by the client.

Client:
435 mm
Factory
+/- 1 mm

8
QUALITY: of a product…
A product has good quality when:
it fits with the requirements specified by the client.

Client:
435 mm
Factory
+/- 1 mm

Restaurant approx. 40 cm

9
AnalyHcal tesHng: your product is a result.

What is a good quality result?

10
Analytical testing: your product is a result.

What is a good quality result?

“A good result
provides a reliable information,
within an agreed precision and accuracy.”

11
Good Q. result = reliable information,
within an agreed precision and accuracy.

12
Good Q. result = reliable informa8on,
within an agreed precision and accuracy.
Quality Management (or Good Lab. PracSces)

Quality Assurance (QA)

Quality Control (QC)

13
Good Q. result = reliable information,
within an agreed precision and accuracy.
Quality Management (or Good Lab. Practices)
Represents all activities which objective is the production
of good quality results.
Quality Assurance (QA)

Quality Control (QC)

14
Good Q. result = reliable informa8on,
within an agreed precision and accuracy.
Quality Management (or Good Lab. PracSces)
Represents all activities which objective is the production
of good quality results.
Quality Assurance (QA)
All actions made to provide confidence in quality of
your result.
(The Quality Assurance Officer should be someone independent of the work.
If no QA officer is available, then Laboratory manager (=YOU!) performs this job.)

Quality Control (QC)

15
Good Q. result = reliable informa8on,
within an agreed precision and accuracy.
Quality Management (or Good Lab. Practices)
Represents all acSviSes which objecSve is the producSon
of good quality results.
Quality Assurance (QA)
All actions made to provide confidence in quality of
your result.
(The Quality Assurance Officer should be someone independent of the work.
If no QA officer is available, then Laboratory manager (=YOU!) performs this job.)

Quality Control (QC)

The techniques and acSviSes that are used to saSsfy
quality requirements.
16
DEFINITIONS & CONCEPTS
(continued)

RESULT:

17
DEFINITIONS & CONCEPTS
(continued)

RESULT:
TRUE VALUE, MISTAKE, ERROR, UNCERTAINTY,
PRECISION, ACCURACY....

18
For analyHcal tesHng:
impossible to make an exact measurement
==> impossible to get the true value.
For analyHcal tesHng:
impossible to make an exact measurement
==> impossible to get the true value.

A good result will only be

as close as possible
from the true value.

20
For analyHcal tesHng:
impossible to make an exact measurement
==> impossible to get the true value.

A good result will only be

as close as possible
from the true value.

Why impossible to reach the true value?

21
For analytical testing:
impossible to make an exact measurement
==> impossible to get the true value.

A good result will only be

as close as possible
from the true value.

Why impossible to reach the true value?

because you will always have errors.

22
True value: possible if
you are coun6ng.

23
True value: possible if
you are counting.

24
True value: possible if
you are counting.

Number of eggs = 3 x 4 = 12

25
True value: possible if
you are coun6ng.

Number of eggs = 3 x 4 = 12
Finding 10 or 11 is a mistake,
mistake ≠ error

26
True value: possible if
you are coun6ng.

Number of eggs = 3 x 4 = 12
Finding 10 or 11 is a mistake,
mistake ≠ error

but measuring is different from counting!

27
measurement has imperfec6ons

28
measurement has imperfections

==>measurement give rise to errors

29
measurement has imperfections

==>measurement give rise to errors

measures ==> errors

30
Error (some exemples):

31
Error:

32
Error:

≠ 39.7
≠ 39.8

33
Error:

≠ 39.7
≠ 39.8

≠ 39.71?
≠ 39. 72?
anyway my
result will
have an error

34
Error:

the difference between

your result and the "true" value.
Error:

36
Error (one more exemple...):

1.4 !

37
Error:

1.4 !

38
Error:

1.4 !

39
Error:

1.4 ! ≠ 1.4 !

40
Error (the most common one):

41
Your objective:

to get as close as possible from the

true value,

that you can NEVER reach…

42
To describe the quality of analyses:

Precision: the closeness of the replicates.

Accuracy: the closeness to the "true" value.

First Mee3ng of the GLOSOLAN, FAO

43
Headquarters, Rome, 1-2 Nov 2017
True value = middle of the target..

44
If you are a beginner:
If you are a beginner:

random distribution!
large errors
after some training:
after some training:
you can improve, having arrows together
(presicion)
after some training:
you can improve, having arrows together
but still not in the center, (not accurate)
after more trainings:
PRECISE and ACCURATE!
after more trainings:
PRECISE and ACCURATE!

t i o n s
r a t u l a
co n g
but if you stop training…
random again….
random again….
To describe the quality of analyses:

Precision: the closeness of the replicates.

Accuracy: the closeness to the "true" value.

First Mee3ng of the GLOSOLAN, FAO

55
Headquarters, Rome, 1-2 Nov 2017
Precision: high low

Accuracy:
high

low
Precision: high low

Accuracy:
high

low
Precision: high low

Accuracy:
high

low
Precision: high low

Accuracy:
high

low
Precision: high low

Accuracy:
high

low
Precision: high low

Accuracy:
high

low
Precision: high low

Accuracy:
high

low
Accuracy and precision in every day life….
Accuracy and precision in every day life….

going through a door with:

high precision
but low accuracy
Accuracy and precision in every day life….
Accuracy and precision in every day life….
Accuracy and precision in every day life….
Accuracy and precision in every day life….
Accuracy and precision with machines….
Accuracy and precision with machines….
very nbecessary in this case...
Error:

the difference between your

result and the "true" value.

Where is the error coming from?

Analytical errors can be of 2 categories:
AnalyCcal errors can be of 2 categories:

1.Random or ‘unpredictable’ deviaSons between

replicates
Analytical errors can be of 2 categories:

1.Random or ‘unpredictable’ deviations between

replicates
Analytical errors can be of 2 categories:

1.Random or ‘unpredictable’ deviations between

replicates

average,
consensus value
AnalyCcal errors can be of 2 categories:

2. SystemaCc or ‘predictable’, regular deviaSon

from the "true" value.
AnalyCcal errors can be of 2 categories:
2. SystemaSc or ‘predictable’, regular deviaSon from
the "true" value.
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.
Analytical errors can be of 2 categories:

2. Systematic or ‘predictable’, regular deviation

from the "true" value.
AnalyCcal errors can be of 2 categories:

2. SystemaCc or ‘predictable’, regular deviaSon

from the "true" value.

100.?
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.

In France: 20°C
100 mL
AnalyCcal errors can be of 2 categories:
2. SystemaSc or ‘predictable’, regular deviaSon from
the "true" value.

In France: 20°C
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.

In France: 20°C
100 mL
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.
In Bhutan: 5°C (glass
shrinkage)

100 mL
AnalyCcal errors can be of 2 categories:
2. SystemaSc or ‘predictable’, regular deviaSon from
the "true" value.
In Bhutan: 5°C (glass
shrinkage)
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.
In Bhutan: 5°C (glass
shrinkage)

100 mL
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.
In Bhutan: 5°C (glass
shrinkage)
In Lao PDR: 35°C (glass
expansion) 100 mL
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.
In Bhutan: 5°C (glass
shrinkage)
In Lao PDR: 35°C (glass
expansion) 100 mL
Analytical errors can be of 2 categories:
2. Systematic or ‘predictable’, regular deviation from
the "true" value.
In Bhutan: 5°C (glass
shrinkage)
In Lao PDR: 35°C (glass
expansion) 100 mL
Analytical errors can be of 2 categories:
2. Systematic (or ‘predictable’ regular deviation from
the "true" value), quantified as "mean difference" (i.e.
the difference between the true value and the mean of
replicate determinations).
Analytical errors can be of 2 categories:
2. Systematic (or ‘predictable’ regular deviation from
the "true" value), quantified as "mean difference" (i.e.
the difference between the true value and the mean of
replicate determinations).
AnalyCcal errors can be of 2 categories:
2. SystemaSc or ‘predictable’, regular deviaSon from
the "true" value.

BIA
S
First Meeting of the GLOSOLAN, FAO
93
Headquarters, Rome, 1-2 Nov 2017
First Mee3ng of the GLOSOLAN, FAO
94
Headquarters, Rome, 1-2 Nov 2017
Analytical errors can be of 2 categories:
1.Random (or ‘unpredictable’ deviations between
replicates); must be made as small as possible.

2. Systematic (or ‘predictable’ regular deviation from

the "true" value); miust be detected and eliminated.
Error:

it is an idealised concept:
errors cannot be known exactly.
you must always keep in mind:
you must always keep in mind:

Despite all efforts,

you cannot avoid
errors.
you must always keep in mind:

Despite all efforts,

you cannot avoid
errors.

= you will always

make errors !
You will always make errors.

Good Laboratory Practices,

will limit the errors.
You will always make errors.

Good Laboratory Practices,

will limit the errors.
Quality Control (QC) help you
to detect the remaining errors.
II – Description of

GOOD LAB PRACTICES.

102
‘GLP’
= low-cost basic measures that will improve
the performances of your laboratory.
‘GLP’
= low-cost basic measures that will improve
the performances of your laboratory.

⚠
This process can be successful only with:
1. a change in daily aWtudes & prac6ces,
2. a change that all the staff must adopt.
GLP tries to correct ‘old habits’ by providing
written documents for all important actions.

The success of GLP also depends on the

cooperaCon, parCcipaCon, involvement
and contribuCon of all laboratory staff.
Benefits of GLP:

- minimize errors

- improve efficiency (thus reducing costs)

- allow quality control (tracking errors & their origin)

- stimulate and motivate all the staff.

What are the:

Good Laboratory Practices?

(GLP)
What are the:

Good Laboratory PracCces?

(GLP)

the laboratory structure:

nested ac6vi6es.
The laboratory structure:
nested ac6vi6es.
Management

Technical elements
Reagents, instruments, etc..

Analytical tasks

pH, NPK, etc..

AnalyCcal tasks

pH, NPK, etc..

with GLP you need:
STANDARD OPERATING PROCEDURES
(SOP)

The objecSves of a SOP is

to do all the important operaSons:
(i) correctly,
(ii) always in the same way.

111
with GLP you need:
STANDARD OPERATING PROCEDURES
(SOP)
Detailed, written instructions to achieve uniformity of
the performance of a specific function.
Repeated application of unchanged processes and
procedures, and their documentation.

112
with GLP you need:
STANDARD OPERATING PROCEDURES
(SOP)
Detailed, wrifen instrucSons to achieve uniformity of
the performance of a specific funcEon.
Repeated applicaEon of unchanged processes and
procedures, and their documentaSon.

SOP are mandatory instructions!

You have to strictly follow a SOP, not to adapt it.

113
!!!
OP
S
This is also something that we experience in daily life
example of the door

and airplane
116
The laboratory structure:
nested ac6vi6es.

Technical elements
Reagents, instruments, etc..

AnalyCcal tasks
Hood, ovens, etc.

Instruments

Reagents

Samples
Technical elements

1. Primary measuring equipment,

2. Analytical instruments,
3. Miscellaneous equipment,
4. Reagents,
5. Soil samples.
Technical elements
1. Primary measuring equipment,
= pipettes, diluters, burettes, thermometers,
balance, sieves, crushers, etc.
They do not provide analy?cal results, but are
necessary to prepare your samples, reagents,
solu?ons, etc.
They must be clean and calibrated.
They must be used correctly.
Technical elements vieux
pHmetre
pourri
2. Analytical instruments
= pHmeter, spectrophotometers, etc.
Technical elements
2. Analytical instruments
= pHmeter, spectrophotometers, etc.
For all instruments, you must have:
- an ‘Operation Instruction Manual’,
-a ‘Maintenance Logbook’.
Technical elements
2. Analytical instruments
= pHmeter, spectrophotometers, etc.
For all instruments, you must have:
- an ‘Operation Instruction Manual’,
-a ‘Maintenance Logbook’.
Instruments should be:
1. suitably located and adequate capacity,
2. periodically inspected, cleaned, maintained,
and calibrated according to SOP.
Technical elements
3. Various equipment & materials
= ovens, fridges, pumps, stills, glassware, etc.
Technical elements
3. Various equipment & materials
= ovens, fridges, pumps, s?lls, glassware, etc.
For all apparatus, you must have:
- an ‘Opera8on Instruc8on Manual’,
-a ‘Maintenance Logbook’.
Technical elements
3. Various equipment & materials
= ovens, fridges, pumps, stills, glassware, etc.
For all apparatus, you must have:
- an ‘Operation Instruction Manual’,
-a ‘Maintenance Logbook’.

Apparatus should be:

1. suitably for the job,
2. properly organized, well cared for.
Technical elements
4. Reagents
Technical elements
4. Reagents

One of the most important sources of the

errors is: using
wrongly prepared or old reagents.
Technical elements
4. Reagents

One of the most important sources of the

errors is: using
wrongly prepared or old reagents.
Reagents have to be :
-prepared very carefully,
-well labelled with: preparation date, expiry dates,
operator’s name.
Record preparations in a Reagents Book.
Technical elements
5. Soil samples
Technical elements
5. Soil samples

Need of proper packaging, labelling and

man agement of samples before going to
the laboratory.
Technical elements
5. Soil samples

Need of proper packaging, labelling and

man agement of samples before going to
the laboratory.
Avoid
- samples being accidentally interchanged,
- being contaminated (broken bags),
- losing their iden?ty (i.e. their label or number)
- or geRng lost.
The laboratory structure:
nested ac6vi6es.
Management

Technical elements

Analytical tasks
Quality control in the soil laboratory:
a statistical process
to monitor and evaluate
the analytical process & results’ quality.

134
III –
INTERNAL & EXTERNAL
QUALITY CONTROL (QC).

135
Thanks for your a-en/on

First Meeting of the GLOSOLAN, FAO

136
Headquarters, Rome, 1-2 Nov 2017