close

Вход

Log in using OpenID

embedDownload
By Saranpat Khaongamdechawat
July 2, 2014
1
2
QUALITY TEST RESULT
LABORATORY TESTING
Quality Result
Analytical process
Staff
Reagent
3
Instrument
4
INTERNAL QUALITY CONTROL (IQC)
Monitor
กก
Error (Low detection error)
!ก"กก#$ Mean '() SD
Leaflet ./"01
Action of detecting error
6
5
Quality Control Situation
SD '2!ก"3ก143(4ก !ก" False rejection 14#8!9:!(
'ก.'()!ก"2#$:!;"<
.=>3
2SD
Mean ± 2SD
3SD
Repeat
SD limit
Too narrow
Minimum Quality
Maximum cost
Buy new control
material
High Pfr
7
8
QC chart control by traditional 2SD
ก QC ก
Negative Trend
Negative shift
9
10
Internal Quality Control (IQC)
ANALYTICAL QC
Requirements for IQC Chemistry
Internal Quality Control (IQC)
External Quality Assessment (EQA)
Stable
sample matrix
At least 2 concentration levels
Established mean and SD range
Levey-Jennings Charts
11
12
Internal Quality Control (IQC)
How to set laboratory mean and SD?
Ex. Control 2 Levels
2
levels of quality control sera on 20 consecutive days.
Calculate mean, SD
Set in Levey-Jennings Charts
13
In The Real !!!!
14
Establishing QC ranges using CVh
Mean ± 2SD
Repeat
Current lot.
SD limit
Too narrow
High Pfr
Buy
new control
Repeat
QC
materialTest
Recalibrate
Buy new QC material
Test Glucose
QC Lot. 45652
Mean = 120.7
SD = 3.18
%CVh = 2.6
N = 553
Comparing your CVh with Manufacturer’s CVs and Peer CVs (EQA)
%CVh < Manufacturer
%CVh < %CV EQA
15
16
Adopt in IQC Program
Establishing QC ranges using CVh
New establishing QC ranges :
Mean = 125.6
SD = 3.26
Test Glucose
QC Lot. 45652
Current lot.
%CVh = 2.6
QC Lot. 45662
Mean = 125.6
N = 20
New lot.
SD = CVh /100 X Mean
= 2.6 / 100 X 125.6
SDNew = 3.26
17
18
Multi--Rules Chart
Multi
In or Out of Control ???????
1. Single rule
Ex. LeveyLevey-Jennings chart with control limits set
MEAN +/+/- 2SD
MEAN +/+/- 3SD
2. MultiMulti-Rules QC
uses 5 different control rules to judge the
acceptability of an analytical run.
19
20
Random Error
Warning Rule
12s
13s
The patient test results in
this run can be reported.
?(ก14 IQC
./22C
2CDE;19<
#$)!9:3>3 (single
control material) !
H</?(
3>3/:I3/ก mean ± 3 SD
8
:=ก ?(ก14 IQC ./
22C 2C D E;19< #$ ) !9 : 3> 3
(single control material) !
H</?(3>3/:I
3/ก mean ± 2SD
21
22
Random Error
Systematic Error
R4s
2 2s
!
H</?(ก14 IQC ./
22C
2CDE;19<#$)!9:3>3
(single control material) 433 2
2>"/ก3 9?(/:I19<
Mean + 2SD 8H/ Mean N 2SD
23
!
H</?(ก14 IQC ./
22C
2CDE;19<#$)!9:ก3 (Single
control material) 433 2 2> "/ก3 9
?(83=</:I3/ก Mean + 1SD '()/9ก?(
83=</:I3/ก Mean N 1SD 8H/#3(ก0D)19<
14#82 2 2>ก3 4SD
22C
2CDE;19<#$ 2 ) (Two
control material) 2)'ก/:I3/ก
Mean + 1SD '()/9ก)83=</:I3/ก
Mean N 1SD 8H/#3(ก0D)19<14#82 2
2>ก3 4SD
24
Systematic Error
Systematic Error
4 1s
!
H</?(ก14 IQC ./
22C
2CDE;19<#$)!9:ก3
(Single control material) 433 4
2> "/ก3 9?(/:I19< Mean + 1SD
8H/ Mean N 1SD 8H/#$กD919<#$
22C
2CDE;/) (Two
control materials) '();?( IQC
2 2> "/ก3./22C
2CDE;'())
9?(/:I19< Mean +
1SD 8H/ Mean -1SD
10x
!
H</?(ก14 IQC ./
22C
2CDE;19<#$)!9:
3>3 (Single control material)
433 10 2> "/ก3 9?(/:I19<
3#383=<./ Mean 8H/#$
#3กD919<#$22C
2CDE;/
) (Two control materials) '()
;?( IQC 5 2>"/ก3./
22C 2C D E;' ( )) 9
?(/:I19<3#383=<./ Mean
25
Systematic Error
26
Systematic Error
2 of 32s เปนกฎที่ใชแสดงการ ปฏิเสธ เมื่อผลการทํา IQC
ของสารควบคุ ม คุ ณ ภาพที่ ใ ช ส ามระดั บ (three
control
materials) จํานวน 2 ใน 3 ครั้ง มีผลอยูที่ mean + 2SD หรือ
mean – 2SD
6x เปนกฎที่ใชแสดงการ ปฏิเสธ เมื่อผลการทํา IQC ของสาร
ควบคุมคุณภาพที่ใชสามระดับ (three control material)
จํานวน 6 ครั้ง ติดตอกัน มีผลอยูที่ดานใดดานหนึ่งของ mean
N=3
N=3
31s เปนกฎที่ใชแสดงการ ปฏิเสธ เมื่อผลการทํา IQC ของสาร
ควบคุมคุณภาพที่ใชสามระดับ (three control materials)
จํานวน 3 ครั้ง ติดตอกัน มีผลอยูที่ mean + 1SD หรือ mean
– 1SD
27
9x เปนกฎที่ใชแสดงการ ปฏิเสธ เมื่อผลการทํา IQC ของสาร
ควบคุมคุณภาพที่ใชสามระดับ (three control material)
จํานวน 9 ครั้ง ติดตอกัน มีผลอยูที่ดานใดดานหนึ่งของ mean
28
กก
ก IQC (QC Records)
Summary Rules Chart
•
- Warning
•
•
12s
•
- Systemic error
•
•
22s , 2 of 32s , 31s ,41s, 6x , 9x, 10x
10x
•
- Random error
•
•
13s , R4s
•
ก
ก mean , SD ( %CV ก
กก,- Performance 3ก
(45 ( Turn-around
time
3ก IQC 3(3 >,?,
3ก
44@ out of control
กกBC4กD?(ก3
Control material (BC4กD?
E F ก-@,กG
>Hก,BC4,?I(ก
ก
3ก
ก Calibrate Calibrator ?IFIกK (กกKL>
29
30
Lab
Result
External Quality Assessment (EQA)
EQA: SC
31
32
EXTERNAL QUALITY ASSESSMENT (EQA)
EXTERNAL QUALITY ASSESSMENT (EQA)
Requirements for EQA
Registration with EQA scheme
organizers
Different serum levels to cover
patient range
stable and consistent sample matrix
matrix mimics patient samples
samples assayed ‘blind’
frequent analysis
rapid feedback
user friendly reports
Objective : Provides laboratories to monitor
1. Long
Long--term precision and accuracy
2. Compare performance with other laboratories
laboratories.
EQA
EQA
Accuracy
33
EQA:SC
34
EQA:SC
VIS = [BIS]
BIS = ( X – DV ) x 100 x 100
DV
CCV
0 - 50
51 - 100
101-150
> 151
A
B
C
Sus.Q
X = Your result
DV = Designed Value(Mean)
CCV = กก IEQAS UK NEQAS
35
36
RIQAS
EQAH--Thyroid
EQAH
37
38
QUALITY REQUIREMENT
Quality
requirement ก
!!ก "#$%&& six sigma * Allowable total error (TEa)
TEa
345 error 6789:;<&=8$ >?:34 error 9:ก!8
@กกA!%BC (Analytical) 95<E
39
TEa CLIA
Routine Chemistry
Test or Analyte
Alanine aminotransferase
Albumin
Alkaline phosphatase
Amylase
Aspartate aminotransferase (AST)
Bilirubin, total
Calcium, total
Chloride
Cholesterol, total
Cholesterol, high dens. lipoprotein
Creatine kinase
Creatine kinase isoenzymes
Creatinine
Glucose
Acceptable Performance
Target value ± 20%
Target value ± 10%
Target value ± 30%
Target value ± 30%
Target value ± 20%
Target value ± 0.4 mg/dL or ± 20% (greater)
Target value ± 1.0 mg/dL
Target value ± 5%
Target value ± 10%
Target value ± 30%
Target value ± 30%
MB elevated (present or absent) or Target value ± 3 SD
Creatinine
Target value ± 0.3 mg/dL or ± 15% (greater)
Target value ± 6 mg/dL
mg/dL or ± 10% (greater)
ก*ก"#$ TEa F&& Absolute concentration limit B* Percent
Example : Glucose
TEa = 6.0 mg/dL
TEa = 10%
GLUC 10 %
-ก,- 100 %
6.0 mg/dL -ก,- (6.0/10) x 100 = 60 mg/dL
(
ก 60 mg/dL IF TEa = 6.0 mg/dL)
https://www.westgard.com/clia.htm
QC Specification
42
QC Procedure ก
Probability
of error detection (Ped) > 0.90
Probability of false rejection (P
(Pfr) < 0.05
QC Rules
QC
Procedure
Number of control (N)
Number of run (R)
QC PLANNING TOOL
1. Sigma metric
2. Operator specifications chart (OPSpec chart)
3. Commercial program
46
Sigma metric calculation
Sigma metric calculation
Sigma metric = [(TEa Bias)/SD]
Sigma metric = [(TEa %bias)/%CV]
47
48
Example
Glucose TEa = 6 mg/dL, 10%
Concentration cut off = (6/10 x 100) = 60 mg/dL
Statistic or Run
Date
QC Procedure by Rule of thumb
Glucose
L1: 45641
L2: 45642
L3: 45643
Current (Mean)
59.7
116.2
346.2
Current (SD)
0.84
1.53
4.49
Current (CV)
1.41
1.32
1.30
Bias
-1.43
%Bias
2.59
5 Sigma
QC Procedure
≥6
IF N= 2 4 N = 3 ( control limit 3.0s 4 3.5s (
ก 2S >I4` (45
Pfr 5
IF N = 2 4 N = 3 ( Control limit 2.5s 4 3.0s ก 3.0s Sigma > 5 (ก
2.5s Sigma < 5
IF N = 4 4 N=6 ( Control limit multirule (13s/22s/R4s/41s) 4 2.5s >KIF multi-rule ก4,>I4
Ped > D?
4
<4
Sigma
(6-1.43)/0.84
(10-2.59)/1.32
(10-2.59)/1.30
Sigma
5.5
5.6
5.7
<3
IF maximum QC procedure non-statistical QC procedures F
preventive maintenance g functional check I4 D? performance
validation check (IF3j-,
FCกIกก
(45
ก
(45I44>,kI4D?ก
49
$@Jก<8 Sigma metric
Christian H.H. Schoenmakers, Andre J.M. Naus, Henricus J. Vermeer, Douwe van Loon and Gerard Steen
1. !"#$%&'(#&)*+ QC Specification (Ped ./0 Pfr) )*+1234ก&%
126(%&078 9:;+<&"&%=>?!>@A&ก Sigma metric graph $4;<%@&+):BCA&ก
ก&%DE@FG%.ก%"(*"H216I*%8<J&6%KA%?G$4;"4AJ&7C#&L
2. Analytical quality requirement $4;DE@A0!>@"&A&ก TEa 6$#&CSBC
3. "4 Rules ./0 N D7@6/V*กC@*L
4. W?@DE@I@*+"4)@*"?/./0$SกX0 6HV;*$4;A0G%06"2C method performance
characteristics ./0I@*+"4(?#"V*ก&%DE@+&C./0I@*+.Lก!1@ES>6AC
52
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Max
Min
Max-Min
Mean
SD
CV
TEa
%Bias
Sigma
CREm
Lab C
1.002
1.0
1.093
1.1
1.044
1.0
1.079
1.1
1.031
1.0
1.082
1.1
1.031
1.0
1.048
1.0
1.074
1.1
1.045
1.0
1.063
1.1
1.038
1.0
1.033
1.0
1.028
1.0
1.081
1.1
1.011
1.0
1.037
1.0
0.991
1.0
1.051
1.1
1.028
1.0
1.093
1.1
0.991
1
0.102
0.1
1.04
1.04
0.03
0.05
2.64
4.73
0.3
0.3
0.08
0.08
7.8
4.4
ก QC Planning ?
$6 IQC %8<&9: 1 กA!%BC Creatinine
$6 precision 9:@789N!; 1 JFB5
13s / 2of3 2s / R4s / 31s
Sigma metric = 4.4
$69:=8$@ก@789N!; 2 JFB5
13S
Sigma metric = 7.8
54
Quality control planning
Normalized OPSpec Chart
OPSpec chart (8
(8 Steps)
1. Quality Requirement
Define quality required for
diagnostic test
Evaluate rejection
characteristics
Assess method
performance
Select control rules,
number of measurements
Prepare OPSpec chart
Adopt Total QC Strategy
Plot operating point
Reassess for changes
3.
-Rejection
characteristics
for difference
control rules
(Pfr)
-Number
control of
measurement
(N)
-Number of
run (R)
2. Precision and Accuracy observed for a method
Normalization
Normalization
กF35 %Imprecision F% %Inaccuracy
A!WกJAX <A;5 Test Glucose TEa = 10 %
34กF35
%Imprecision F% %Inaccuracy 9;&
ก<&5 TEa F$AF35ก343C>PC
Q*:9:RJ5=3 Plot " OPSpecs Chart U8;
*ก OPSpecs Chart 9:B%9:78JB<& Test
<EV
Normalized Imprecision
สูตร
= %CV/TEa x 100
Normalized Inaccuracy
สูตร
= %Bias/TEa x 100
57
58
Normalization
ตัวอยาง Test Glucose มี TEa = 10 %
Normalized Imprecision
สูตร
= %CV/TEa x 100
= 1.59/10 x 100
= 15.9 ( แกน X )
Normalized Inaccuracy
สูตร
= %Bias/TEa x 100
= 0.19/10 x 100
= 1.9 ( แกน Y )
59
60
Normalize OPSpec Chart (N=2 and N=4)
Normalize OPSpec Chart (N=3 and N=6)
N=3 and
90% AQA
N=6 and
90% AQA
N=3 and
50% AQA
N=6 and
50% AQA
61
Normalize OPSpec Chart
62
Normalize OPSpec Chart
63
64
Normalize OPSpec Chart
Normalize OPSpec Chart
65
66
Normalize OPSpec Chart
Normalize OPSpec Chart
50% AQA, N=3
90% AQA, N=3
67
Normalize OPSpec Chart
Normalize OPSpec Chart
50% AQA, N=6
90% AQA, N=6
Normalized Imprecision
= 15.9 ( แกน X )
Normalization
Normalized Inaccuracy
= 1.9 ( แกน Y )
ตัวอยาง Test Glucose มี TEa = 10 %
Normalized Imprecision
สูตร
= %CV/TEa x 100
= 1.59/10 x 100
= 15.9 ( แกน X )
Normalized Inaccuracy
สูตร
(15.
15.9 , 1.9)
= %Bias/TEa x 100
= 0.19/10 x 100
= 1.9 ( แกน Y )
71
72
เลือก Control Rules
โดยเลือกจากเสนที่อยูเหนือ Operating Point และ False Rejection นอยๆ
(ไมเกิน 5% หรือ 0.05)
(15.
15.9 , 1.9)
73
3%U;#C"ก9J QC planning
8A:;5ก;
\กA!%BC\8! Q8
ก@`5 total quality control strategy (TQC strategy) -LO-Ped strategy (Low error detection) rIF maximum QC ,
FK,- error ก 13S /2 of 3 2S/R4S/31S/6X N=6
I4
,Cก non statistic QC procedure I4กD? ก
กIF optimize QC process stability, กIF3j-,F C
( กIF patient data QC
Quality control
planning
&53]^B
9:ก!8?E=8$R6ก$ A8PA
8&3%X5"#$@5;
"ก9J quality control
#5A;Q!:ก9J quality control
=8$;53%!9W!_Q
77
1/--pages
Пожаловаться на содержимое документа