BASIC QUALITY CONTROL STATISTICS Flashcards
The mean (or average) is the laboratory’s best estimate of the
analyte’s true value for a specific level of control.
Mean
Standard deviation quantifies how close numerical values are
in relation to each other
Standard deviation
implies that data points are close to the mean
Low SD
implies that data points are spread out over a
large range of values
High SD
A graph of all the quality control results in a month
Levey Jennings Chart
In levey jenning → in a month, minimum of ____controls should be run
20
→ concentrations
Y axis
line in the middle
Mean
→ days or number of run
X axis
→ exactly ___ runs of control in a day (per shift)
3
__________ is a measure of variability
Coefficient of variation
A method or instrument’s _______ is expressed as a percent
Coefficient of variation
A statistic that measures your lab’s bias relative to your
consensus group. An estimate of reliability
Standard deviation index
→ Ideal score; Identical to peer group
→ data values of the lab are in 100% agreement with the
reference method
SDI - 0
→ acceptable to marginal performance
→ may need to investigate test system bias
SDI -1 -1.5
→ marginal performance
→ may need to perform corrective action
1.5-2.0
A statistic that compares your lab’s precision to that of other
labs in a consensus group.
Coefficient variation of ratio
→ Acceptable performance
→ Indicates that precision is better than peer group
Less than 1 - CVR
→ Acceptable to marginal performance
→ May need to investigates test system imprecision
1 to 1.5 - CVR
→ Marginal performance
→ May need to perform corrective action
1.5-2.0 - CVR
Is defined as abrupt changes in the control values
Shift
→ if there is 6 or more consecutive QC results already fall in
the one side of the mean → reject results
Shift
● Shifts in QC data represent a sudden and dramatic positive or
negative change in test system performance
Shift
Sudden failure or change in the light source
Shift
Change in reagent formulation
Shift
Change of reagent lot
Shift
Major instrument maintenance
Shift
Sudden change in incubation temperature (enzymes only)
Shift
● Change in room temp. or humidity
Shift
Failure in sampling system & reagent dispense system
Shift
Inaccurate calibration/recalibration
Shift
Indicated a gradual loss of reliability in the test system
Trend
→ if there is 6 or more consecutive QC results that slowly but
continually move in one direction → data values are seen
on one side of the mean → reject results
Trend
Deterioration of the instrument light source
Trend
Gradual accumulation of debris in sample/reagent tubing
Trend
Gradual accumulation of debris on electrode surfaces
Trend
Aging of reagents
Trend
Gradual deterioration of control materials
Trend
Gradual deterioration of incubator chamber temp.
Trend
Gradual deterioration of light filter integrity
Trend
Gradual deterioration of calibration
Trend
→ a trend or shift away from the laboratory mean
Systematic error
→ not acceptable; indicates that there are errors in the
system
Systematic error
any random deviation away from the laboratory mean
random error
variations on QC results that shows no pattern
random error
does not reflect failure on some parts of the testing system
random error
→ not likely to reoccur
random error
→ rejected if the data value that showed the random error
exceeds ± 2SD
random error
If 2 levels of controls
Westgard multirule system
● a “multi-rule” system developed by _________
based on statistical concepts
Dr. James O. Westgard
a combination of decision criteria or rules to assess if a
system is in-control
Westgard multirule system
● use when at least 2 levels of control are run with the
examination run
Westgard multirule system
● cannot use with only one control
Westgard multirule system
This is a warning rule that is violated when a single control
observation is outside the ± 2s limits.
12s rule
Two consecutive controls exceed the mean by +/- 2SD
22s rule
4 consecutive controls on one side of the mean exceeds
+/-1SD
41s rule
10 consecutive controls on one side of the mean
10x rule
Improper alignment of sample or reagent pipettes resulting in
change in reagent and calibrator volume.
Systematic error
● Drift or shift in incubator chamber temperature, deterioration
of photometric light source
Systematic error
● Change of calibration/reagent lot
Systematic error
● Deterioration of reagent/calibrator/control product while in
use, storage or shipment
Systematic error
● Incorrect handling of control product
Systematic error
Inadequate storage of reagent or calibrators/control products
Systematic error
Change in test operator
Systematic error
Any QC result outside ± 3s violates this rule.
13s rule
If there is at least a 4s difference between control values
within a single run, the rule is violated for random error.
R4s rule
Power Supply
Random error
Double pipetting of control sample
Random error
Misplacement of Control sample within the run
Random error
Random air bubbles in reagent or sample pipette system
Random error
Incorrect reconstitution of the control product
Random error
Inappropriate Storage of control
Random error
Inadequately mixed reagents
Random error
Individual operator variation in pipetting , timing , etc.
RAndom error