Chapter 4: Analytical and

Post-analytical QA

Part 1 . Quality Control Samples,

Charts and Rules
Learning Objectives

 Identify different types of quality control

 List elements found in Levey-Jennings charts and Shewart
graphs
 Calculate quality control data
 Use Levey-Jennings charts
 Interpret quality control data using common quality control
rules.
Learning Objectives …

Upon completion of this lecture the student should be able

to:
 Describe a quality control sample in terms of assayed
versus unassayed, internal and external and their use.
 Indicate the main components of a Levey-Jennings chart
including values and markings on the x and y axis.
Learning Objectives …

 Describe the following QC rules in terms of appearance with

control values and their ability to detect systemic (shifts and
trends) versus random precision errors:
1
3s
2
2s
R
4s
6
x
7
t
4
1s
Chapter Outline

 Analytical and Post-analytical Quality Assurance

 The Need for Quality Control
 Types of quality controls
 Quality Control Materials
 Calibrators
 Quality Control Use
 Levey-Jennings or the Shewhart control charts
 Quality Control Rules
Outline …

Approaches used to interpret patient samples in quality control

 Absurd value check
 Duplicate analysis
 Delta check
 Mixed samples
 False sense of security
 Post Analytical Quality Assurance
 Documentation of tests results

 Interpretation and Reporting of tests results

4.1 The Need for Quality Control
(QC)
 Analytical phase of QA is quality control
 Patient results are unknowns.
 Analysis produces a number.
 Is the result valid?
 Patient may have previous result.
 Change from previous may be expected
 No change from previous result may be expected

 Quality control samples are known.

The Need for QC …

Quality control is used to test the analytical phase of patient

testing.
 It is a process or system for monitoring the quality of
laboratory testing, and the accuracy and precision of
results
 Quality control samples are tested along with patient
samples to monitor the validity of the analysis.
The Need for QC …

 If quality control samples which have a known

concentration do not give expected results, it can often
mean an error occurred that affected patient results as
well.
Purpose of Quality control:
 Quality Control is used to monitor both the precision
and the accuracy of the assay in order to provide
reliable results
4.2 Types of quality controls

1. Internal quality controls

 The internal QC involves the in-house procedures for
continuous monitoring of operations & systematic day-to-
day checking
 To decide whether the data are reliable enough to be
released.
 The procedures primarily monitor the bias of data with the
help of control samples
 The precision by means of duplicate analyses of test
samples and/or of control samples.
Types of quality controls …

 IQC specimens comprises

These activities take place at batch level (second-line
control).
 in-house patient sera (single or pooled)

 international serum standard

They should be interpreted according to the criteria given
by the manufacturer.
Types of internal quality
control materials
I. Aqueous based control mater
 accurately measured reagent grade chemicals + purified water +
some other appropriate solvent.
 less expensive than other types
 stable for long periods of time
 require no reconstitution or mixing
 difference in the physical matrix
 lower viscosity , lighter color, and less turbidity .
 affect the way the material responds
 these possibly reducing its ability to detect SE .
II. Lyophilized controls

 are protein based control

 prepared by dry freezing system.
 made of human serum pool
 The pool is well mixed and dispensed in to sterile vials.
 The preservatives, antimicrobial agents and specific
chemicals added on it
change its viscosity
Lyophilized controls …

 large pools are prepared at one time

 making a large volume of a single lot number of the
product available for distribution.
 unreconstituted lyophilized controls will remain stable for
months even for years.
Lyophilized controls…

 The final material has the same properties as fresh

human serum.
 Reconstitution of the material can be a major source of
error
 Prone to vial to vial variations
 May not available for immediate use
III. liquid serum controls

 Serum pools diluted with ethylene glycol (30%)

 Ethylene glycol lowers the freezing point
 be stored at low as -20°c & still remain liquid
 stabilizes the sample and also acts as an antimicrobial agent
 well mixed, dispensed in to plastic squeeze bottles and
stored in the freezer.
liquid serum controls …

 There is no preparation, waiting time & problem of

turbidity and the shelf-life after reconstitution is long.
 ethylene glycol increases the viscosity
 This is a significant factor if the material is to be used in
the test system that requires
 the specimen to be either dialyzed
liquid serum controls….

 pumped through very small diameter pump tubing

 to be spread evenly over the surface such as a dry
chemistry reagent strip or slide.
 it interferes with freezing point osmolality determinations

and more expensive.

Types of internal quality
control materials …
2. External controls
It involves reference help from
 other laboratories
 participation in national and/or international interlab
sample
 data exchange programmes (proficiency testing).
 They are used to monitor consistent performance, lot to
lot variation b/n kits and Indicator of assay performance
on samples
Types of internal quality
control materials …

 Ideal set of external controls should include

-Positive controls

-Negative controls

- Border line reactor

 If the laboratories choose to use only one external
control, border line reactor must be given priority
4.3 Quality Control Materials

 Similar to patient specimens

 Preserved for consistency
 Assayed –provided with mean and SD for analytes
 Unassayed – you must determine mean and SD
 QC materials may be made in-house
 QC materials may be made by manufacturer
 The analyte concentration should be at medically
significant level
Internal QC materials: Patient
replicates

 Previously analysed patient sample

 Easily available
 Cost effective
 Results and samples readily available
 stable for the period of use
4.4 Calibrators

 Contain known concentration of the analyte dissolved in a

simple matrix
 Use to adjust the output of an instrument to analyte
concentration (calibration)
 Calibration occurs when first setting up an instrument.
 Calibration occurs periodically
 according to manufacturer recommendations

 according to guidelines written in operating procedure

 Similar to standard
Difference Between a
Calibrator and Control?

 Both have known concentrations of the analyte(s) specified

 Calibrator is used to adjust the output of an instrument
before measuring unknowns like patient samples.
 Calibrator is run less frequently than a patient sample,
 Control is used as a “known patient” to test the validity of
the analytical run, eg. Daily
4.5 Quality Control Use
 Qualitative testing of QC materials:
 Provide same type of result as patient specimens
 Positive (weak) or Negative

 Recorded daily on monthly log but not plotted on a

QC chart
 Semi-quantitative testing of QC materials:
 include the analyte concentration at each of the grade
level that is negative, trace, 1+, 2+, 3+---
 each control level can be used as comparison
 tested daily and with the opening of new bottle
Quality Control Use …

 Quantitative testing of QC materials:

 Analyzed solely for quality control purposes
 Not for calibration
 checked by using a low to high range of two or three
controls
 Plotted on monthly charts to track over time
Preparing Quality Control
Materials
 External: follow manufacturer’s instructions
 Dilute
or reconstitute with diluents
 Mix well and store according to instructions

 Internal: detailed instructions

 Poolsalvaged
 Determine mean and SD
 Preserve for use in next run
Qualitative QC Results

Month: Year:
Lot # Positive Control
Lot # Negative Control
Lot # Kit Exp. Date:
Day Positive Negative Technologist
control control
4.6 Quantitative QC Results
and Expectations
 Gaussian distribution
Quantitative QC Results …

Control rule:
Is a rule or a decision criterion for judging weather an
analytical run is in control or out of control.

 It is commonly defined by a symbol of AL, were A is a

control measurement and L identifies the control limits
 Commonly mean +/- SD.1SD, 2SD, 3SD are used
QC Rules are Used as an Alarm
System
 QC result should
 indicate a true alarm
 Not indicate a false alarm
QC Outside of Expected Range

 QC value > +3SD or < -3SD from the mean

 Commonly indicates a problem
 QC value > + 2SD or < -2 SD from the mean
 5% could be due to acceptable result
 False alarm if reject without further checks
 QC value > +1 SD or < -1 SD from the mean
 Expected most of the time
Historical Use of QC (Mean
Chart)

 Levey and Jennings – quality standards for the

laboratory
 Shewart developed statistical tools to monitor quality
 Others
Shewart Chart

 Shewart set up separate

mean and standard
deviation or range charts to
evaluate quality.
 Mean of daily QC results are
plotted.
 S of daily QC results are
plotted on a separate chart.
Levey-Jennings Chart
 Levey-Jennings Chart
3 SD

2 SD

1 SD
Mean

1 SD

2 SD

3 SD

LJ chart with QC samples

Levey-Jennings Chart...

 It is the graphical method of displaying control results

 evaluating weather procedures is with in or out of
control.
 The L_J chart indicates.
Mean (x)
Standard deviation (SD) +1, +2, +3---
95% confidence limits for the following results
 On the Y-axis the data or number of analysis and on the
X-axis the value of control.
Levey-Jennings Chart …

 Mean (X) = The sum of the values (X1+X2+X3…X31)

divided by the number (n) of observation
 Variance=the square of SD=(Xi-X)2
n-1
 Coefficient of variation (CV) = SD X 100%
Mean(X)
 Lower 95% Conf Limit =Mean (x)-2SD (S)
 Upper 95% Conf Limit =Mean (x) +2SD(S)
Recording of quality control
results on Levey-Jennings
quality control charts
Quality Control Limits

 Mean is the central line.

 One standard deviation above = + 1 s
 One standard dev below = - 1s
 2 standard dev above = + 2s
 2 standard dev below = -2s
 3 standard dev above = + 3s
 3 standard dev below = -3s
Calculating Confidence Limits
on the Chart
 For example, Control 1 has a mean of 200 and a
standard deviation of 4 mg/dL.
 What is the 95% confidence limit or mean + 2 s limit?
 The upper control limit would be:

200 + 2*4 = 208 mg/dL

 The lower control limit would be:

200 - 2*4 = 192 mg/dL

4.7 Quality Control Rules

 Decision Criteria
 In control versus out of control
 control limits
 Mean + 2s
 95% limits
 5% false rejection
QC with each analytical run

 How long is a run?

 Manual
 Automated method
QC Procedure

 Quality control SOP

 How many controls
 How often run
 How documented
 How problems are solved
Quality Control Rules

 Should detect random errors

 Imprecision

 Should detect systematic errors

 Shift (6 results that settle around new mean)
 Trend (7 results upward or downward )
QC Rules

 Accept analytical run if both QC results fall within + 2 SD

of the mean if no shifts or trends are present.
 Acceptable Ranges : + 2SD of the mean

Instrument: _________________ Test: ____________________ Units: ________

Method: ____________________ Month/Year: ______________ Tech: ________
Control: ____________________ Lot #: ____________________ Exp: _________
Control Mean: _______________ 1 Std Dev: ________________ -+3 SD Range: __________

+ 3s(90)

+ 2 s (85)

+ 1 s (80)

Mean (75 mg/dL)

- 1 s (70)

- 2 s (65)

- 3 s (60)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Days
13s rule violation

A run is rejected when a single control measurement

exceeds the mean plus 3s or the mean minus 3s control
limit.
12s rule violation

This rule is used as a warning rule to trigger careful inspection

of the control data by the following rejection rules
22s rule violation

Reject when 2 consecutive control measurements exceed

the mean plus 2s or minus 2s control limit.
QC Rules…….

 The QC result fell outside of 2 s on second consecutive

day or run
 Systematic Error detected with 2 2s
 This means the patient results are held/ rejected
R4s rule violation

Reject when 1 control measurement in a group exceeds the

mean plus 2s and another exceeds the mean minus 2s.
41s rule violation

Reject when 4 consecutive control measurements exceed

the same mean plus 1s or the same mean minus 1s
control limit
6X Rules

 The 6th consecutive day the same control result was

lined up with the previous days.
 Systematic Error detected with 6 x: shift
 This means the patient results are held/ rejected
7t Rules

 The 7th consecutive day this control result fell in an

upward or downward pattern.
 Systematic error detected with 7t: trend
 This means the patient results are held/ rejected
10x rule violation

Reject when 10 consecutive control measurements fall on

one side of the mean
 Modifications of 10x to make it fit more
easily with N‘s of 4

 When N is 4, that could mean 2 measurements on each

of two different control materials, or 4 measurements on
one material, or 1 measurement on each of four
materials
8x rule violation

Reject when 8 consecutive control measurements fall on

one side of the mean.
12x rule violation

Reject when 12 consecutive control measurements fall on

one side of the mean.
QC Multirule System

 These rules work best in a multi-rule system

 Rules evaluated in a certain order
 Warning rules trigger the need to look for other violations

1 2s
Westguard and his associates

 used when two levels of control material are analyzed

per run
 high concentration and the other low concentration

The most frequently used control rules include?

 12 SD rule, 13 SD rule,
 22 SD rule, R4 SD rule, 41SD rule
 10X rule
Westgard Multirules

 This decision tree help to decide whether to reject or

accept a run. If you answer yes, it means that rule has
been violated. 4 1s may be replaced by 7t and 10x may
be replaced by 6x.
Westgard Multirules …

 Reject analytical run if

 One control result exceeds the mean ± 2SD for a
second time within 20 days. Before the patient results
reported.
 Both control results exceed the mean ± 2SD
 A shift or trend occurs
Summary Part: 1

 Control materials can be described in terms of

 assayed (mean and s provided)
 unassayed (mean and s not provided)
 Use (validation of run not as a calibrator)
Summary Part :1…

 Levey-Jennings chart including values and markings on

the x (days or time) and y axis (concentration mean and
s). Errors such as shifts and trends can be detected.
Summary Part: 1 …

 Plotting QC over time and applying QC rules can help to

detect random errors in patient results.
The principle of a Control Chart of the Mean. UCL =
Upper Control Limit (or Upper Action Limit). LCL = Lower
Control Limit (or Lower Action Limit). UWL = Upper
Warning Limit. LWL = Lower Warning Limit.
Part 2

Approaches used to interpret patient

samples in quality control
Learning Objectives

Upon completion of this lecture the student should be able to:

 Differentiate quality assurance measures such as absurd
value check, duplicate analysis and delta check.
 Discuss the use of mixed samples
 Describe the false sense of security in laboratory analysis
 Post Analytical Quality Assurance
 Describe correct documentation, interpretation and
reporting of tests results
Lecture Outline

 Absurd value check

 Duplicate analysis
 Delta check
 Mixed samples
 False sense of security
 Post Analytical Quality Assurance
 Documentation of tests results

 Interpretation and Reporting of tests results

4.8 APPROACHES USED TO
INTERPRET PATIENT SAMPLES IN
QUALITY CONTROL

 Absurd value check

 Duplicate analysis
 Delta check
 Mixed sample
 Average of normal
Absurd value check

 Biological Limit check

 Examples of results not compatible with life
 Sodium result of 100 mmol/L

 Hemoglobin result of 2 g/dL

 Glucose result of 2 mg/dL

 Internal quality assurance measure

 These results indicate the need to check the
instrument, reagent or specimen for errors and retest
specimen.
Duplicate analysis

 Test 10 consecutive patient specimens in duplicate

 Standard deviation (s)
2 s is the limit for difference between duplicates
 Useful as internal quality assurance measure
Delta Check

 Compare patient parameters with previous run

parameters from the same patient
 If a difference considered an analytical error
 precision and accuracy
 Example
 Hemoglobin and MCV
 Na
 Internal quality assurance measure
Mixed samples

 It is the old and seldom used method and as follows.

 Equal portion of two randomly picked samples are mixed
and analyzed
 The initial results of the individual samples are averaged
to determine a target value for the mixed sample. The
difference is plotted
 Twice the average difference is used as a control limit
False sense of security

 Controls within range feeling of correctness but upon closer

inspection this is incorrect assumption,
 this is because controls are different from patients’ sample in
that control materials are:
 not drawn from patient
 not handled in the same way
 no error during collection
 no effect of centrifugation, evaporation
 no wrong tube drawn
 no transcriptional error
False sense of security …

 Note;- Good performance with controls does not

necessary mean that the over all performance of the
laboratory is good.
 For these reason controls are with in range, does not
necessarily indicate patient value are correct hence the
pt samples are prone for pre-analytical errors
Interpretation and Reporting
Patient Results
 For interpretation of microscopic or gross examination
results
 Expected normal results should be established
 Morphological changes should be determined based
on standardized criteria
 Reference ranges should be established for all
quantitative test results
 For Interpretation of patient quantitative results
 Compare the Patient Result with the Reference Range
Post Analytical Quality
Assurance

 includes calculation, recording, reporting, and verification

of test result
 taking appropriate action whenever a result has a
serious clinical implication
 ensuring test results are interpreted correctly.
Reporting the patient result

 3 Rs
 Reliability

 Rapidity

 Relevance
Reporting of Patient Results

 Use approved form

 Fill in relevant information:
 Patient name and accession number
 Ordering physician
 Specimen collection information
 Patient Result
Reporting of Patient Results …

 Reference Range
 Name of reporting technologist
 Date and time of report
 Comments by reporting technologist
 Reviewer name, date and time
Standardization in reporting
test results

 Results and units

 Interpretation of results
 Benefit of standardized reporting
 Comparison of results
 Efficiency of laboratory service
 Referral of patient from another health facility
Patient Report Form
Reporting of Patient Results …

Results should be carefully review before reporting to clinicians.

Avoid common Post-analytic Errors:

 Wrong name
 Incorrect units or reference range
 Transcribing wrong number
 Report too late
Document Examples

Record book
Specimen receipt log
Documentation and Record
Management
 Manage patient test records in whatever method works
best
 Carbon copies of reports
 Worksheets
 Register or logbooks
 Manage test records in worksheets or registers
 Work reports
 Laboratory workload
 Instrument maintenance
 Quality control logs
 Occurrence logs
Documentation

 Document storage and retrieval

 Use format that works best
Summary Part 2

 Absurd value check, duplicate analysis and delta

checks are used for post-analytical quality assurance
before patient results are reported.
 Mixed samples
 A false sense of security in laboratory analysis
 Correct documentation in specimen and result
registers, interpretation of patient results compared to
reference ranges and reporting of patient tests results
on the approved test report form are important for post-
analytical quality assurance.
Review quastions

1. What is the use of patient samples as quality control

material perceived as attractive?
 2. Two laboratory technologists determine the hemoglobin
values of 120 patients using hemocue methods. The
following table showed the individual results.

LAB
Personnel I
Anemic Anemic Non anemic Total

LAB Anemic 40 60 100

Personnel II
Nonanemic 50 50 100
Total 90 110 200
What is the kappa value of the two readers to
diagnose anemia using hemocue methods.
Show all the steps to calculate the kappa values
and interpret the level of anemia in the two
readers/ laboratory technologists.
References

1. Shewhart WA. Economic Control of Quality of

Manufactured Product. New York; D. Van Hostrand
Company, Inc., 1931.
2. Levey S, Jennings ER. The use of control charts in the
clinical laboratory. Am J Clin Pathol 1950;20:1059-66.
3. Henry RJ, Segalove M. The running of standards in clinical
chemistry and the use of the control chart. J Clin Pathol
1952;27:493-501.
4. Westgard JO, Groth T, Aronsson T, Falk H, deVerdier C-H.
Performance characteristics of rules for internal quality
control: probabilities for false rejection and error detection.
Clin Chem 1977;23:1857-67.
References …

5. Westgard JO, Barry PL, Hunt MR, Groth T. A multi-rule

Shewhart chart for quality control in clinical chemistry. Clin
Chem 1981;27:493-501.

6. Westgard JO, Barry PL. Cost-Effective Quality Control:

Managing the Quality and Productivity of Analytical
Processes. Washington, DC:AACC Press

7. www.westgard.com