STATISTICAL CONCEPTS in QUALITY CONTROL Flashcards
A statement of the extent of random variation in any series of measurement
STANDARD DEVIATION
A measure of the distribution of values around the mean
STANDARD DEVIATION
Square of the standard deviation
VARIANCE
Used to detect significant differences between groups of data
VARIANCE
Determine contributions of various factors to the total variation
VARIANCE
Percentile expression of the mean
COEFFICIENT OF VARIATION
Measure of the relative magnitude of variability
COEFFICIENT OF VARIATION
CV = s (100)/X
COEFFICIENT OF VARIATION
Reported values do not fall along the line of slope when graphed
Analytical Bias
Reported lab results do not correspond to the correct value
Analytical Bias
Problem on Accuracy (Bias) = ____________
Systematic error
Refers to laboratory analyses that are subject to imprecision
Random Analytical Variability
Problem on Precision (CV) = _____________
Random Error
can be systematic or personnel-related
Errors
Sources cannot be completely controlled or identified
RANDOM ERRORS
Increase the extent of variability of results
RANDOM ERRORS
Displace the mean value on one direction (may be up or down), but do NOT affect the overall variability as shown by the SD value
SYSTEMATIC ERRORS
decomposition of reagents during storage
Aging phenomena
Series of values on the control chart that continue to increase or decrease for at least a period of six (6) consecutive days
TREND
Values that distribute themselves on one side of the mean for at least six (6) consecutive days
SHIFT
Q.C. procedures performed on a daily basis within individual laboratories
INTRALABORATORY / INTERNAL
Monitoring is carried out using:
o Levey – Jennings Chart
o Westgard Multi-rule Chart
o CUSUM Technique
Performed on a less frequent basis (e.g. 3x a year) to compare performance b/w or among laboratories
INTERLABORATORY / EXTERNAL
graphical representations that display the control observation as a function or time
QUALITY CONTROL CHARTS
Control results are plotted on the Y-axis (ordinate) vs time on the X-axis (abscissa)
LEVEY – JENNINGS CHART
a control value that goes beyond ±2SD
Outlier
Uses a series of control rules for interpreting data
False rejection is kept low
- WESTGARD MULTIRULE CHART
Error detection is improved
Analyse samples of the control material by the analytical method to be evaluated (at least 20 days)
WESTGARD MULTIRULE CHART
seeks to identify, measure, and eliminate the large gaps of inefficiency in a process
SIX SIGMA
Provides a display of the differences b/w the observed values & the expected mean
CUMULATIVE SUM (CUSUM) CHART
Used to monitor small shifts in the process mean.
CUMULATIVE SUM (CUSUM) CHART
Obtain the CUSUM by adding the difference (value from the mean) to the CUSUM of the previous
differences
CUMULATIVE SUM (CUSUM) CHART
Under the interpretation od CUSUM chart. If it is a steep slope, it suggests that?
presence of SE (Out of Control)
found in clinical Laboratories with microcomputer Q.C. programs
DECISION LIMIT CUSUM
Laboratory’s observed mean for material A (Y-axis) compared to the observed mean for material B
(X-axis)
YOUDEN PLOT / SCATTER DIAGRAM
Information about the nature of the SE can be obtained when two diff. control materials have been analyzed
YOUDEN PLOT / SCATTER DIAGRAM
Presents information in an efficient, graphical format
CHECK SHEET
Accomplished with a simple listing of items
CHECK SHEET
Used to identify factors that have the greatest cumulative effect on the system
PARETO CHART
Used to associate multiple possible causes with a single effect
- Ishikawa or Fish bone diagram
Pictorial representations of a process
Flow chart
- Used in identifying where errors are likely to be found in the system
Flow chart
Provides a simple, graphical view of accumulated data, including its dispersion & central tendency
HISTOGRAM
Ease of construction
HISTOGRAM
