WS/T 641-2018 PDF in English
WS/T 641-2018 (WS/T641-2018, WST 641-2018, WST641-2018)
| Standard ID | Contents [version] | USD | STEP2 | [PDF] delivered in | Name of Chinese Standard | Status |
| WS/T 641-2018 | English | 245 |
Add to Cart
|
0-9 seconds. Auto-delivery.
|
Internal quality control for quantitative measurement in clinical laboratory
| Valid |
Standards related to (historical): WS/T 641-2018
PDF Preview
WS/T 641-2018: PDF in English (WST 641-2018) WS/T 641-2018
HEALTH INDUSTRY STANDARD OF
THE PEOPLE’S REPUBLIC OF CHINA
ICS 11.020
C 50
Internal quality control for quantitative
measurement in clinical laboratory
ISSUED ON: DECEMBER 11, 2018
IMPLEMENTED ON: JUNE 01, 2019
Issued by: National Health and Wellness Committee of the People's
Republic.
Table of Contents
Foreword ... 3
1 Scope ... 4
2 Terms and definitions ... 4
3 Preparation before internal quality control carried out ... 5
4 Design of internal quality control method ... 6
5 Actual operation of internal quality control ... 11
6 Management of internal quality control data ... 17
7 Quality control method of applicable patient data ... 18
8 Comparison between laboratories for internal quality control data ... 19
Annex A (informative) Common quality control rules and meanings ... 21
Annex B (informative) Power function diagram method ... 23
Annex C (informative) Operational process specifications diagram method .. 25
Bibliography ... 27
Internal quality control for quantitative
measurement in clinical laboratory
1 Scope
This Standard specifies the purpose of internal quality control for quantitative
measurement items in clinical laboratory, the design of internal quality control
method, the actual operation of internal quality control, the management of
internal quality control data, the quality control method based on patient data
as well as the comparison of internal quality control data between laboratories.
This Standard is applicable to the quantitative measurement in clinical
laboratory of the medical institution that carries out the clinical inspection
service.
2 Terms and definitions
For the purposes of this document, the following terms and definitions apply.
2.1 quality control
a part of quality management that is committed to meeting quality requirements
[GB/T 19000-2016, 3.2.10]
2.2 internal quality control
the inspector, according to a certain frequency, continuously measure the
specific components in a stable sample; use a series of methods to analyze;
according to statistical law, infer and evaluate the reliability of the measurement
results for this batch so as to determine whether the inspection report can be
sent out; timely discover and exclude the dissatisfaction factors in quality link
2.3 quality control strategy
quality control variety, testing frequency of each variety, position of placement,
as well as rules for interpretation of quality control data and rules for
determination whether the analysis batch is under control or out of control
2.4 measurement bias [JJF 1001 5.5]
bias for short
3.3 Calibration
The instrument used to measure clinical sample shall be calibrated according
to some certain requirements. In calibration, select suitable (supporting)
calibration product. If possible, ensure that test results are traceable to
reference methods or/and reference materials. For different analysis items, it
shall, according to the characteristics, establish their own calibration
frequencies.
3.4 Quality control product
3.4.1 Characteristics
The quality control product shall have the similar or same matrix with the testing
sample of patient. The quality control product shall be uniform and stable. If
conditions permit, store the quality control product that can be used for one year
or more. The inter-bottle variability shall be less than the variation of analysis
system. If there is no commercialized quality control product, the laboratory can
make the quality control product by itself.
3.4.2 Analyte concentration in quality control product
The concentration of selected quality control product shall be within the
concentration range that is clinically meaningful.
If the fixed-value quality control product is used, the original calibration value in
the user’s manual can be only used as reference. It must be repeatedly
measured by the laboratory to determine the tentative and common mean value
as well as the standard deviation.
4 Design of internal quality control method
4.1 Quality control method selection and design form
The quality control selection form is a 3×3 form. It determines quality control
method that is applicable to nine different types of inspection procedures
(quality control rules and number of quality control results per batch N). For
single-rule fixed-limit quality control method, establish quality control selection
and design form, for example, Levey-Jennings quality control chart. For multi-
rule quality control method, establish quality control selection and design form,
for example, multi-rule quality control method. Table 1 and Table 2 respectively
show two quality control selection and design forms. The process capability of
the form row is described by medically important systematic error (ΔSEc). The
process stability of the form column is described by frequency of errors, f.
new batch number shall be determined with the currently used quality control
product. According to the results of at least 20 quality control measurements
obtained by 20 or more independent batches (excluding abnormal values or
outlying values), calculate average value as tentative mean value.
Use this tentative mean value as the central line of next month’s internal quality
control diagram for internal quality control. After one month, gather the under-
control results of this month and the results of previous 20 quality control
measurements. Calculate the accumulated average (the first month). Use this
accumulated average as the average value of next month’s quality control
diagram.
Repeat the above operation process. Continue three to five months or
continuously accumulate on a monthly basis.
5.1.1.2 Setting of common mean value
Use the accumulated average number that is calculated by the first 20 data and
all data collected by 3~5 months’ under-control data AS the common mean
value of quality control product within validity period. And use it as the average
number for future internal quality control diagram. For the individual item that
the concentration level continuously changes within validity period, the average
value needs adjusting constantly.
5.1.2 Quality control product of short stability
Within 3~4 days, analyze 3~4 bottles per level of quality control products every
day. Repeat 2~3 times per bottle. After data collection, calculate the average
number, standard deviation and coefficient of variation. Inspect the abnormal
value to data. If an abnormal value is found, it needs to recalculate the average
number and standard deviation of the rest data. Use this mean value as the
central line of quality control diagram.
5.2 Setting of control limit
For quality control product that has new batch, it shall determine the control
limit. The control limit is usually represented by the standard deviation multiple.
5.2.1 Quality control product of long stability
5.2.1.1 Setting of tentative standard deviation
To determine the standard deviation, the quality control product that has new
batch number shall be tested together with the currently used quality control
product. According to the results of at least 20 quality control measurements
obtained by 20 or more independent batches (excluding abnormal values or
outlying values), calculate the standard deviation and use it as tentative
5.3.2 Frequency of quality control product inspection
Within each analysis batch, at least perform one inspection for quality control
product. The manufacturer of inspection system or reagent shall recommend
the amount of quality control product for each analysis batch. User, according
to different situation, may increase or reduce the number of quality control
measurements.
5.3.3 Position of quality control product
User shall determine the positions of each batch of internal quality control
products. The principle is that before reporting the inspection results of a batch
of patients, it shall evaluate the quality control results. When determining the
position of quality control product, it must consider the type of analysis method
and possible error type. For example, within the user specified batch length,
perform non-continuous sample inspection. If the quality control product is put
before the end of the specimen inspection, it may monitor the bias. If the quality
control product is evenly distributed within the entire batch, it may monitor the
drift. If it is randomly inserted in the patient sample, it may inspect the random
error. In any case, it shall evaluate the quality control results before reporting
patient inspection results.
5.3.4 Replacement of quality control product
When it plans to replace the quality control product with new batch number, it
shall be performed before the end of the “old” batch number of quality control
product. Measure the new batch number of quality control product and “old”
batch number of quality control product together. Repeat the processes of 5.1
and 5.2. Establish new mean value and control limit.
5.4 Drawing of quality control diagram and record of quality control
results
The quality control diagram is to measure and record the process quality, so as
to assess and monitor if the process is under control. It is a diagram of a
statistical method design. There are central line (CL), upper control limit (UCL)
and lower control limit (LCL) on the diagram. There shall also be quality control
results that are in chronological order or trace sequence of quality control
results’ statistical value. According to the mean value and control limit of quality
control product, draw Levey-Jenning quality control diagram (single
concentration level), or Z-score diagram that different concentration levels are
drawn on the same diagram or Youden diagram. Record the original quality
control results on the quality control chart. Reserve the printed or electronic
original quality control record.
5.5 Application of quality control rules
there shall be corresponding measures to verify the patient inspection results.
6 Management of internal quality control data
6.1 Statistical processing for internal quality control data per month or
within specified time
At the end of each month, after the last batch of test results or within specified
time, it shall collect and statistically process all quality control data. The
calculation shall at least include:
(1) average number, standard deviation and coefficient of variation of each
measurement item’s raw quality control data in current month or specified
time;
(2) average number, standard deviation and coefficient of variation of each
measurement item, after excluding out-of-control data in current month or
specified time;
(3) accumulated average number, standard deviation and coefficient of
variation of all quality control data of each measurement item in current
month or specified time as well as the past, after excluding out-of-control
data in current month or specified time.
6.2 Storage of internal quality control data per month or within specified
time
At the end of each month or within specified time, it shall collect and organize
all quality control data and keep in the archives. The quality control data for
archive shall include:
(1) raw quality control data of all items in current month or specified time;
(2) quality control diagram of all items’ quality control data in current month
or specified time;
(3) all calculated data in 6.1 (including average number, standard deviation,
coefficient of variation and accumulated average number, standard
deviation, coefficient of variation);
(4) out-of-control report in current month or specified time (including the out-
of-control rule it is against, out-of-control reason, correction measures
adopted).
6.3 Quality control data chart reported per month or in specified time
At the end of each month or within specified time, after collecting and organizing
For a specific patient, if the situation is stable, the test results before and after
of the patient shall be basically stable. Therefore, when the patient’s situation
is stable, the difference between patient’s continuous test results, that is
Δ(delta), shall be small. If the Δ value is large and exceeds the pre-defined limit,
then it means that one of the following three possibilities shall exist: (1) the test
results of patient sample do change; (2) specimen is marked wrong or mixed;
(3) one of the two calculated results of Δ value has error.
Usually, it uses one of the following two methods to calculate Δ value:
Δ (experimental unit) = the second result – the first result
Δ (%) = 100 × (the second result – the first result) / the second result
7.3 Extreme deviation quality control diagram method of double-testing
of patient sample
Some inspection procedures use double-testing. At this time, when using the
difference of double-testing value of patient sample, it can determine the in-
batch standard deviation of inspection procedure. It can also use the extreme
deviation of double-testing to inspect the in-batch random error.
The difference of double-testing value can be drawn on the Shewhart's extreme
deviation quality control chart. Its quality control limit can be calculated from the
difference’s standard deviation. It can also use the following formulas to export
the control limit of double-testing extreme deviation from double-testing
standard deviation (sdouble):
R0.025 control limit = sdouble × 3.17
R0.01 control limit = sdouble × 3.64
R0.001 control limit = sdouble × 4.65
8 Comparison between laboratories for internal quality
control data
8.1 Comparison between laboratories for internal quality control data of
quality control product
If many laboratories share a same batch of quality control products, it may
organize the report results as a comparison plan between laboratories.
Obtain statistical data from the data of this plan. Use it to determine:
(1) the imprecision in laboratory and between laboratories;
Annex A
(informative)
Common quality control rules and meanings
The quality control rules are used to explain quality control measurement
results and determine the state of analysis batch control. They are represented
by the symbol AL, where, A is the number of quality control measurement results
that exceed the control limit (L) or the statistics of quality control measurement
results, and L is the control limit. When the quality control measurement results
meet the conditions required by the rules, it shall determine that this analysis
batch violates these rules.
Common quality control rules are: ( refers to average number; s refers to
standard deviation)
12s: means that one quality control measurement result exceeds ; if it
violates this rule, it shall prompt warning;
12.5s: means that one quality control measurement result exceeds ; if
it violates this rule, it shall prompt that random error exists;
13s: means that one quality control measurement result exceeds ; if it
violates this rule, it shall prompt that random error exists;
R4s: means that the difference value of two quality control measurement results
of same batch exceeds 4s, that is, one quality control measurement results
exceeds , the other quality control measurement results exceeds ;
if it violates this rule, it shall mean that random error exists;
22s: means that two continuous quality control measurement results exceed
and at the same time; if it violates this rule, it shall mean that
systematic error exists;
41s: means that continuous four measurement results of one quality control
product all exceed or , continuous two measurement results of
Annex B
(informative)
Power function diagram method
B.1 Determination of allowable total error
Currently, it may adopt national quality evaluation standard between clinical
laboratories used by Ministry of Health Clinical Laboratory Center, national
health and wellness commission industry standard or allowable total error (TEa)
exported according to biological variation.
B.2 Evaluation of inspection procedures
According to the evaluation plan of inspection procedures, evaluate the
performance parameters that are quantitatively measured by this laboratory
one by one. Determine the imprecision of each item (represented by CV%) and
the bias (represented by bias%).
B.3 Calculation of critical systematic error
Critical systematic error
B.4 Drawing of power function diagram
The power function diagram describes the statistical "power" of quality control
method, where, y-axis is the error detection probability Ped, and x-axis is the
critical error size. Figure B.1 is the power function diagram of critical systematic
error of different quality control rules.
Annex C
(informative)
Operational process specifications diagram method
The diagram of operational process specifications (OPSpecs) is a lining
drawing. It displays the imprecision, the bias, the quality control method
required to the use of inspection procedures. It describes the statistical quality
control method that shall be used so as to reach the allowable imprecision and
bias. It also ensures the probability required by expected quality through routine
operation. The OPSpecs diagram can be used to verify if the current statistical
quality control method is appropriate, or to verify if the new selected quality
method can meet the analysis quality requirements. Since it is unnecessary to
calculate the critical error, and unnecessary operation has been reduced, the
application of OPSpecs diagram can simplify the design process of quality
control method.
Figure C.1 is OPSpecs schematic. This diagram can ensure that 90% of
measurement results do not exceed 10% of allowable total error. In this
OPSpecs diagram, the y-axis is the allowable bias (bias%), and the x-axis is
allowable imprecision (CV%). The uppermost diagonal line (solid line) in the
diagram refers to the maximum allowable limit of imprecision and bias. The
specified total error is (bias)+ 2s. This total error is usually used to determine
whether this method performance is an acceptable standard during method
evaluation. The lower diagonal lines (dashed lines) respectively refer to routine
operation limits when the measurement methods are unstable, when there are
systematic errors, different quality control methods shall be required to perform
quality control (each diagonal line represents a quality control method). When
using OPSpecs diagram, draw the imprecision and bias of the measurement
method on the diagram. Determine the operation point of laboratory. Then
compare it with the routine operation limit of different quality control method.
The quality control method that routine operation limit is higher than operation
point can be adopted. They can achieve the required level of quality assurance
and become an optional quality control method. But the final selection of quality
control method still needs to consider the number of quality control
measurement results required, the out-of-control probability and the difficulty of
implementation.
According to the long-term internal quality control measurement results of each
clinical laboratory, it may estimate the fixed imprecision or random error (CV%)
of inspection procedures. According to the confirmation or verification or
correctness of inspection procedures, verify the quality evaluation plan between
laboratories to obtain the bias (bias%).
Bibliography
[1] Westgard JO, Barry PL, Hunt MR. A multirule Shewart chart for quality
control in clinical chemistry. Clin Chem. 1981; 27:493-501.
[2] Westgard JO, Barry PL. Cost-Effective Quality Control: Managing the Quality
and Productivity of Analytical Processes. Washington, DC: AACC Press;
1986:230.
[3] Hyltoft Petersen P, Ricos C, Stöckl D, et al. Proposed guidelines for the
internal quality control of analytical results in the medical laboratory. Eur J Clin
Chem Biochem. 1996; 34:983-999.
[4] Westgard JO. Internal quality control: planning and implementation
strategies. Ann Clin Biochem. 2003;40(Pt 6):593-611.
[5] Westgard JO. Six Sigma Quality Design & Control: Desirable Precision and
Requisite QC for Laboratory Measurement Processes. 2nd ed. Madison, WI:
Westgard QC, Inc.; 2006.
[6] CLSI C24A3. Statistical quality control for quantitative measurement
procedures: principles and definitions. 3rd Ed. Wayne (PA): Clinical and
Laboratory Standards Institute; 2006.
[7] CLSI C24-Ed4. Statistical quality control for quantitative measurement
procedures: principles and definitions. 4th Ed. Wayne (PA): Clinical and
Laboratory Standards Institute; 2016.
[8] Westgard JO. Basic QC Practices, 4th ed. Madison, WI: Westgard QC, Inc.,
2016.
[9] Editor-in-Chief Yang Zhenhua. Clinical laboratory management (for medical
testing). Beijing: People's Health Publishing House, 2003.
[10] Editor-in-Chief Shen Ziyu. Hospital management - Clinical laboratory
management. Beijing: People's Health Publishing House, 2003.
[11] Editor-in-Chief Li Ping. Clinical laboratory management (national college
medical planning textbook). Beijing: Higher Education Press, 2006.
[12] Editors-in-Chief Shen Ziyu, Li Ping. Clinical laboratory management (2nd
Edition). Beijing: People's Health Publishing House, 2007.
[13] Editor-in-Chief Chen Wenxiang. Hospital management - Clinical laboratory
management volume (2nd Edition). Beijing: People's Health Publishing House,
2011.
......
Source: Above contents are excerpted from the PDF -- translated/reviewed by: www.chinesestandard.net / Wayne Zheng et al.
|