Quality control and Quality assurance Flashcards
General definition of quality
General definition of quality - Value of services
The Product produced by clinical labs is INFORMATION
data or results
Why monitor quality in a Clinical Lab
- Ensure that results delivered are valid and the information is of value and is reliable
- If the QC results are off, what does it say for the patient results?
- Compliance with CLIA Standards and regulations
Quality levels
QC–> QA–> TQM
Standard( calibrator)
- A material of known value used to standardize (calibrate) any analytical
procedure and thereby determine the value of an unknown (patient) - Use for verification/validation
Primary standards
- Primary
- A solution generally containing a single purified solute (e.g. glucose) in a pure
solvent (e.g. water).
Secondary standards
- Secondary
- A material with an established value(s) determined by assay against a
primary standard. It may contain one or more constituents.
Serum calibrator
- contains numerous constituents (solutes)
- uses serum as the solvent
- solutes are of equivalent purity to those used in primary standards
- solutes added to solvent (serum) on the basis of analytical weighings.
Controls
(Quality Control)
* A stable substance which contains one or more known constituents used to
verify the accuracy and precision of a method * Use for runs
Shelf life-
compare usage
* Use 1 mL/day
* Control 1: $2/mL, 100 mL, 1 month shelf life * Control 2: $5/mL, 100 mL, 6 month shelf life
Box pricing
- Compare price per mL, not per box
Ideal:
Three levels
* Low level (abnormal) * Normal level
* High level (abnormal)
Linearity of instrument
- Controls must read within the linear range of the instrument
TSH Immunoassay
which one is correct
Immunoassay Control Set A
* Low Level (1.03 – 1.23 μIU/mL)
* Normal Level (7.5 – 9.6 μIU/mL)
* High Abnormal Level (27.9 – 34.5 μIU/mL)
Immunoassay Control Set B
* Low Level (3.0 – 5.0 μIU/mL)
* Normal Level (8.0 – 10.0
μIU/mL)
* High Abnormal Level (45 – 55 μIU/mL)
Lab considerations
Laboratory Considerations:
1. TSH Decision Limits: Low (<3 μIU/mL), Normal (3-10
μIU/mL), High (>10 μIU/mL
2. Linearity of Instrument: Linear to 50 μIU/mL
Levy Jennings Chart
- If a control point falls outside the +/- 2SD of the mean value, the control results are considered to be “out of control” and patient samples cannot be run or reported.
- NOTE: ONE control point out of 20 outside of +/- 2 SD is acceptable (95% confidence limits)
- If the control values falls within the +/- 2SD of the mean value, the QC results are “in control” and patient results may be reported
Westgard multi- rules
- set of rules based on a statistical analysis of controls
- System of criteria to determine if a system is “in control”
- Can be used with one control (within) or more than one control (across)
Multirules
- Simply: More than one rule
- Example:
- My diet allowance:
- Onedonutperday
- No more than 3 donuts in a week
- Cannot have 1⁄2 of a donut more than 5 days in a row
West guard rules
Set of rules to determine if analytical run is “in control” or “out of control”
13s rule
when a single control measurement exceeds the mean plus 3s or the mean minus 3s control limit.
12s
-when control limits are set as mean plus/minus 2s.
- when outside this limit then the machine shuts off.
22s
Reject when 2 consecutive control measurements exceed the same mean plus 2s or the same mean menus 2s control limit
R4s
Reject when 1 control measurement in a group exceeds the mean plus 2s and another exceeds the mean minus 2s
41s
Reject when 4 consecutive control measurements exceed the same mean plus 1s or the same mean minus 1s control limit.
41s rule
The criteria which must be met to violate this rule are
-Four consecutive results
- greater than 1s
- on the same side of the mean
There are two application to the 41s rule
- Within control material (e.g. all Level I
control results) - Across control materials (e.g., Level I, II,
and III control results in combination).
Within control material violations indicate systematic bias in a single area of the method curve while violation of the across control materials application indicates systematic error
over a broader concentration.
Warning rule 12s
- If occurs once in every 20 runs, very important to keep results in QC statistics to maintain same bell curve – if rejected will continually make the bell curve narrower.
- If occurs more than once in every 20 runs then indicates a dispersion problem and imprecision. The result should be rejected from QC statistics.
Additional rules
*Random Error Mandatory rules: 13S and R4S detect
*Systematic Error Mandatory rules: 22S and 10X detect bias problems.
10x rule
Reject when 10 consecutive control measurements fall on one side of the mean.
you will sometimes see some modifications of this last rule to make it fit more easily with Ns of 4: 8x - reject when 8 consecutive control measurements fall on one side of the mean.
8x
reject when 8 consecutive control measurements fall on one side of the mean.
8x
reject when 8 consecutive control measurements fall on one side of the mean.
Rules for using the rules
At least two control levels
- At least two control levels (i.e. normal and abnormal) must be performed with each run of patient samples or on each timed shift.
Rules for using the rules
Test the 12s rule if
- Test the 12s rule. If both control values are within +2 standard deviations, accept the run and report out the patient test results. If either control is outside +2 standard deviations, look at the next multi-rule. Hold the patient test results. Once any multi-rule has been broken, then the run is rejected and further analysis of the multi-rules is not continued. Then, problem solving begins.
Test the 13s rule if
- Test the 13s rule. If either of the control results was outside + 3 standard deviations, then reject the run. Do not report the patient test results.
Tests the 22s rule if
- Test the 22s rule. First, review the control results for the normal and abnormal control. If both controls were greater than +2 or –2 standard deviations in the same direction, then reject the run and do not report out the patient results. Secondly, review the results for the last two normal controls and then the results for the last two abnormal controls. Look for results that are greater than +2 and -2 standard deviations in the same direction. If found, then reject the run. Do not report out the patient results.
Tests the R4s rule if
- Test R4s rule using both the normal and abnormal controls in the same run. If one control is greater than +2 standard deviations and the other is greater than –2 standard deviations, reject the run.
Tests the 41s rule if
- Test the 41s rule for the last three normal controls and then the last three abnormal controls. If the last three normal or abnormal controls are greater than +1 or -1 standard deviation, then reject the run. Examine the last three controls, looking at both the normal and abnormal controls. If the last four controls of normal and abnormal levels were greater than +1 or –1 standard deviation, then reject the run.
Tests the 10X rule if
- Test the 10x rule by looking at the last 10 controls run, whether normal or abnormal. Look to see if the results all fall on one side of the mean or the other side. If found, reject the run.
