7/29/2019 BASIC CONTROL RULES
1/39
BASIC QUALITY CONTROL
STATISTICS
Geromil J. Lara, RMT, MSMT
7/29/2019 BASIC CONTROL RULES
2/39
MEAN
The mean (or average) is the laboratorysbest estimate of the analytes true value for
a specific level of control.x=xi/n
Where:
xi= each data point
n= the number of data points in the set
7/29/2019 BASIC CONTROL RULES
3/39
Glucose QC Values
using Level 1 as control material:
Day 1 4.5
Day 2 4.7
Day 3 4.6Day 4 4.4
Day 5 4.6
_______
22.8 / 5 runs = 4.56 Mean
7/29/2019 BASIC CONTROL RULES
4/39
Standard Deviation
Standard deviation quantifies how closenumerical values are in relation to eachother.
Low standard deviation indicates that all ofthe data points are very close to the samevalue (the mean),
High standard deviation indicates that thedata are spread out over a large range ofvalues.
Also used to set up limits
SD = (Xn-Mean)2n-1
http://en.wikipedia.org/wiki/Meanhttp://en.wikipedia.org/wiki/Mean7/29/2019 BASIC CONTROL RULES
5/39
STANDARD DEVIATION
FORMULA: (MEAN) +/- (SD) (1) FOR 1SD
LOWER LIMIT 4.56 - (0.1)(1) = 4.56 - .1 = 4.46
HIGHER LIMIT 4.56 + (0.1)(1) = 4.56 +.1 = 4.66
7/29/2019 BASIC CONTROL RULES
6/39
Formula :(Mean) +/- (SD) (2) for 2SD
Lower Limit 4.56 - (0.1) (2) =4.56 - .2 = 4.36
Higher Limit 4.56 + (0.1) (2) = 4.56 +.2 = 4.76
7/29/2019 BASIC CONTROL RULES
7/39
Formula :(Mean) +/- (SD) (3) for 3SD
Lower Limit 4.56 - (0.1) (3) = 4.56 - .3= 4.26
Higher Limit 4.56 + (0.1) (3) = 4.56 +.3 = 4.86
7/29/2019 BASIC CONTROL RULES
8/39
LEVEY JENNINGS CHART
4.56
4.76
4.66
4.86
4.46
4.36
4.261 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
7/29/2019 BASIC CONTROL RULES
9/39
Coefficient of Variation
The Coefficient of Variation (CV) is ameasure ofvariability.
A method or instruments CV isexpressed as a percent and iscalculated as:
CV (%)= (Standard Deviation / Mean) (100)
Example: 0.1/ 4.56 (100) = 2.19%
7/29/2019 BASIC CONTROL RULES
10/39
Importance of CV
Coefficient of Variation is useful forcomparison of precision for twodifferent methods or instrument.
It can also be used as a part of theInternal Quality Control systemwhen performing patients precisiontesting.
The lower the CV, the better
7/29/2019 BASIC CONTROL RULES
11/39
SDI ( Standard Deviation
Index)
A statistic that measures your labsbias relative to your consensusgroup. An estimate of reliability.
Example : __100 -98__= 1
2
7/29/2019 BASIC CONTROL RULES
12/39
0 - Ideal score; Identical to your peergroup
(+/-)1 to 1.5
acceptable to marginal performance;may need to investigate test system bias
(+/-)1.5 t0 2.0marginal performance;
may need to perform corrective action.
7/29/2019 BASIC CONTROL RULES
13/39
CVR ( Coefficient of Variation
Ratio)
A statistic that compares your labsprecision to that of other labs in a
consensus group.
Ex. 10 = 2
5
7/29/2019 BASIC CONTROL RULES
14/39
Less than 1
Acceptable performanceIndicates that precision is better than peergroup
1 to 1.5Acceptable to marginal performanceMay need to investigates test system
imprecision
1.5 to 2.0Marginal performance
May need to perform corrective action
7/29/2019 BASIC CONTROL RULES
15/39
Using Levey-Jennings Chart to
Evaluate Run Quality
7/29/2019 BASIC CONTROL RULES
16/39
Shift
is defined asabruptchanges inthe control mean.
Shifts in QC data represent asudden and dramatic positive ornegative change in test system
performance
7/29/2019 BASIC CONTROL RULES
17/39
Example of Shift
Mean
2SD
1SD
3 SD
-1SD
-2SD
-3SD1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
7/29/2019 BASIC CONTROL RULES
18/39
Shifts maybe caused by:
Sudden failure or change in the lightsource
Change in reagent formulation
Change of reagent lot
Major instrument maintenance
Sudden change in incubation temperature(enzymes only)
Change in room temp. or humidity Failure in sampling system & reagent
dispense system
Inaccurate calibration/recalibration
7/29/2019 BASIC CONTROL RULES
19/39
Trend
indicates a gradual loss ofreliability in the test
system.
Trends are usually subtle.
7/29/2019 BASIC CONTROL RULES
20/39
Example of Trend
Mean
2SD
1SD
3 SD
-1SD
-2SD
-3SD1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
7/29/2019 BASIC CONTROL RULES
21/39
Trend may be caused by:
Deterioration of the instrument light source
Gradual accumulation of debris insample/reagent tubing
Gradual accumulation of debris onelectrode surfaces
Aging of reagents
Gradual deterioration of control materials
Gradual deterioration of incubator chambertemp.
Gradual deterioration of light filter integrity
Gradual deterioration of calibration
7/29/2019 BASIC CONTROL RULES
22/39
Trends and Shifts areindicative of a problem
and these should beinvestigated and
corrected!
7/29/2019 BASIC CONTROL RULES
23/39
WESTGARD RULES
7/29/2019 BASIC CONTROL RULES
24/39
ERROR DETECTION
Systematic Errora trend or shift away from the laboratorymean
Random Errorany random deviation away from thelaboratory mean
7/29/2019 BASIC CONTROL RULES
25/39
This is a warning rule that is violated when asingle control observation is outside the 2slimits.This rule merely warns that random error orsystematic error may be present in the testsystem.
7/29/2019 BASIC CONTROL RULES
26/39
Violation indicates
systematic error
Two consecutive controls
exceed the mean by +/- 2SD
7/29/2019 BASIC CONTROL RULES
27/39
This indicatessystematic error 4 consecutive controls on one side of the
mean exceeds +/-1SD
7/29/2019 BASIC CONTROL RULES
28/39
This indicates systematic error 10 consecutive controls on one side of
the mean
7/29/2019 BASIC CONTROL RULES
29/39
Sources of
Systematic Error Improper alignment of sample
or reagent pipettes resulting tochange in reagent and calibratorvolume.
Drift or shift in incubatorchamber temperature,deterioration of photometric
light source
Change of Calibration /reagentlot
7/29/2019 BASIC CONTROL RULES
30/39
Sources of
Systematic Error Deterioration of
reagent/calibrator/controlproduct while in use, storageor shipment
Incorrect handling of controlproduct
Inadequate storage of reagentor calibrators/control products
Change in test operator
7/29/2019 BASIC CONTROL RULES
31/39
This rule identifies random error
Any QC result outside 3s violates this rule.
7/29/2019 BASIC CONTROL RULES
32/39
If there is at least a 4s difference betweencontrol values within a single run, the rule isviolate forrandom error.
7/29/2019 BASIC CONTROL RULES
33/39
Sources of
Random Error Power Supply
Double pipetting of control
sample
Misplacement of Control samplewithin the run
Random air bubbles in reagent orsample pipette system
7/29/2019 BASIC CONTROL RULES
34/39
Sources of
Random Error Incorrect reconstitution of the
control product
Inappropriate Storage ofcontrol
Inadequately mixed reagents Individual operator variation in
pipetting , timing , etc.
7/29/2019 BASIC CONTROL RULES
35/39
BIAS/SDI RESULTS
ACCURACY /RELIABILITY ARE AFFECTED
Mean
2SD
1SD
3 SD
-1SD
-2SD
-3SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
7/29/2019 BASIC CONTROL RULES
36/39
PRECISION/REPEATABILITY/CV/
CVI ARE AFFECTED
Mean
2SD
1SD
3 SD
-1SD
-2SD
-3SD
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
7/29/2019 BASIC CONTROL RULES
37/39
IDEAL CONTROL RUNNING
Mean
2SD
1SD
3 SD
-1SD
-2SD
-3SD
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
7/29/2019 BASIC CONTROL RULES
38/39
QC- THE OUT OF CONTROL PROBLEM
Bad Habit #1 Repeat the controls
Bad Habit # 2 Try a new control
Good Habit #1 Inspect the control chart orRules violated to determine the type of error
Good Habit #2 Relate the type of error topotentialCausesGood Habit # 3 Verify the solution and documentthe remedy
Develop Good Habits- Solve Problems
7/29/2019 BASIC CONTROL RULES
39/39
If you think its hard
it will be