QUALITY ASSURANCE

IN HEMATOLOGY
QUALITY ASSURANCE
• Quality Control system in hematology laboratory is COMPLEX
• Unavailability of weighed standards
Requires:
• Elaborate calibration
• Validation
• Matrix effect examination
• Linearity
• Reference interval examination
• Accuracy, integrity, judgment, timeliness
QUALITY ASSURANCE
QUALITY = ACCURATE = REPRODUCIBLE

RELIABILITY
• Vigilance
• Effort
• QA & QC specialist
QUALITY ASSURANCE
• Comprehensive and systematic process that strives to ensure reliable patient
result
• The practice & process of assessing performance in all steps of the
laboratory testing cycle
• Pre analytical
• Analytical
• Post analytical
To promote accurate and excellent result
“ A rigorous quality assurance is the key feature in ensuring quality results”
COMPONENTS OF QUALITY ASSURANCE
1. PREANALYTICAL VARIABLES
• deals with all aspect affecting the test outcome occurring prior to the
testing procedure
• test order
o user friendly

o adequate patient information

• sample collection
“ the test result is only as good as the quality of the sample”
COMPONENTS OF QUALITY ASSURANCE
1. PREANALYTICAL VARIABLES
• proper patient identification
• properly labelled test tubes
• proper anticoagulant
• proper mixing of sample
• timely delivery to laboratory
• tubes checked for clot
• medications administered to patient
• previous blood transfusions
• intravenous line contamination
COMPONENTS OF QUALITY ASSURANCE
2. ANALYTICAL VARIABLES
• addresses all issues involving the test procedure
• laboratory staff competence
• assay & instrument selection
• assay & instrument validation
o linearity, accuracy, analytical limits (AMR)& specificity
• internal & external quality control
COMPONENTS OF QUALITY ASSURANCE
3. POSTANALYTICAL VARIABLES
• addresses factors that can affect the test outcome and its use after the testing
process
• accurate transcription & filing of results
• content & format of laboratory report
• reference/ therapeutic range
• timeliness in communicating critical values
• turnaround time (TAT)
• patient /physician satisfaction
• physician application of laboratory result
• patient outcome
• cost analysis
QUALITY ASSURANCE
• Laboratory assay utilization
• Physician test ordering pattern
 “pre-pre” analytical variables
• Appropriate application of laboratory assay results
 “post-post” analytical variables
• 17% error
• Laboratory directors develop CLINICAL QUERY SYSTEM
• guide clinicians in laboratory assay selection
STATISTICAL COMPUTATIONS
Statistical computation in a hematology laboratory are used for:
• Plotting quality control charts
• Establishing reference ranges
• Performance of correlation studies
• Evaluation of proficiency surveys
• Analysis of data trends
• Management of financial resources
STATISTICAL COMPUTATIONS
• MEAN
 sum of all results divided by the number of results (at least 20
measurements)
 most reliable measure of the center of distribution
 markedly affected by extreme values
STATISTICAL COMPUTATIONS
• MEDIAN
 Data point that separates the upper half from the lower half of a data set
 Robust expression of central tendency in a skewed distribution
• minimizes the effect of outliers
• MODE
 Data point that appears most often in the sample
 Not a true measure of central tendency
• trimodal
STATISTICAL COMPUTATIONS
• VARIANCE expresses the deviation of each data point from its expected
value

• STANDARD DEVIATION - square root of the sum of the squared

differences of each data point from the mean divided by n - 1 for sample
EXPRESSION OF CENTRAL TENDENCY & DISPERSION

STANDARD DEVIATION
 used for predictable measure of dispersion from the mean in a normal
distribution
 describes the average “distance” of each data point from the mean in a
normal distribution
• confidence limit
• degree of random error
• confidence interval (CI)
EXPRESSION OF CENTRAL TENDENCY & DISPERSION

STANDARD DEVIATION
• Confidence interval (CI)
 Dispersion
 ±2 SD or 95.5% ci
 Expression of random or chance variation
 Range of expected value after re testing
• outliers
• data points over 2 SD from the mean
 The larger the SD of a sample, the greater the deviation from the mean.
EXPRESSION OF CENTRAL TENDENCY & DISPERSION

COEFFICIENT OF VARIATION (CV)

• Normalized expression of SD; expressed as a percentage or decimal fraction

• statistical tool used to compare variation for data with different mean value
• utilized to standardize the SD regardless of the magnitude of analyte
concentration
EXPRESSION OF CENTRAL TENDENCY & DISPERSION

COEFFICIENT OF VARIATION (CV)

 most commonly used measure of dispersion
 the lower the CV, the more precise are the data
 the lower the CV, the more consistent is the assay
 less than 5%
 the higher the CV, the greater the dispersion
EXPRESSION OF CENTRAL TENDENCY & DISPERSION
METHOD VALIDATION
• METHOD VALIDATION is a process that is used to demonstrate the
suitability of an analytical method for an intended purpose.
• New assay, assay modification, instrument-assay combination
INCLUDES:
 proof of accuracy
 precision
 reportable ranges (amr)
 detection of interfering substances
 Results are documented
METHOD VALIDATION
• ACCURACY is the measure of agreement between an assay value & the
theoretical “true value”
• Difficult to establish & maintain
• Comparability
• Current reagent - instrument system vs new reagent - instrument system
• Reagent - instrument system vs previously validated external reference
method
• Primary standard - material of known, fixed composition; pure form; dry mass
• Secondary standard - preserved plasma with known concentration
METHOD VALIDATION
Primary Standards in Hematology Calibrators
• Cyanmethomoglobin • properties closely match the test
• Fibrinogen specimen
o preserved human blood cells
• Factor VIII
suspension
• Protein C o nucleated avian rbc/microlatex
• Antithrombin particles
• von Willebrand factor • assayed using reference methods
Internal Quality Control
• PRECISION – the expression of reproducibility or dispersion about the
mean; assessment of random error, variation
• SD or CV%
• Performed on 3 – 5 calibration specimen
• WITHIN DAY PRECISION
 assay single specimen for 20 consecutive times; uses same reagent &
instrument
• DAY TO DAY PRECISION
 same source specimen & instrument; separate aliquots; 20 assays ; 10
runs on 10 consecutive days
Internal Quality Control
• PRECISION
• Compute the mean, SD & CV%
• CV% documents dispersion or random error
 Laboratory professionals equate the quality of an assay with its CV%
• Within-run – 10% or less
o automated differential provides CV% level of 5% or less (800+ cells)

• Day-to-day run – 30%

o depending on stability complexity of the assay

 The best assay is one that combines the lowest CV% with the greatest
accuracy
Internal Quality Control
Internal Quality Control
• LINEARITY – the ability to generate results proportional to the calculated
concentration of activity of the analyte
• High level calibrators are diluted; each dilution is assayed 2x to 3x for precision
• Dilution must span the AMR
• Specimen outside the linear range are diluted, reassayed & computed by the
dilution factor
• Lower limits are important in PC & coagulation assays
• Difference between <1% & 3% factor VIII activity affects treatment options
• Difference between a PC of 10,000/µL & 5,000/µL affects the decision to treat
with platelet concentrate
Internal Quality Control

• ANALYTICAL MEASUREMENT RANGE

– the upper & lower limit of expected values
• Results that fall below or above the AMR
are never reported
• Accuracy is compromised
Internal Quality Control
• ANALYTICAL SENSITIVITY (LLD)
• Lower limit detection
• Required of local laboratory professional when modifying FDA-
approved assay or developing an LDT
• Computed as 3SDs above the mean of blank assay results
• Prevents false-positive result generated by interference (noise)
Internal Quality Control
• ANALYTICAL SPECIFICITY
• The ability of an assay to distinguish the analyte of interest from
anticipated interfering substances w/in the specimen
• “spike” – specimen with potential interfering substances
• Common interferences
• Hemolysis; lipemia; icterus
• QC manager publish procedures to remedy the presence of interfering
substances that may compromise lab test results
Internal Quality Control
LEVELS OF LABORATORY ASSAY APPROVAL (FDA)
• Approved
• Cleared
• Modified-cleared
 approved or cleared assay modified by laboratory professionals
• Analyte-specific reagent (ASR)
 should not be used for non-cleared applications
• Research use only (RUO)
 used on a trial basis
• Laboratory developed test (“home-brew”)
Internal Quality Control
• DOCUMENTATION & RELIABILITY
• Validation record – stored for 2 years
• Recalibration interval – every 6 months or updates in reagent lot
• Control results
• Instrument repair

 CLINICAL LABORATORY IMPROVEMENT AMENDMENTS of 1988

(CLIA)
Internal Quality Control
• LOT-TO-LOT COMPARISON
• Changes is no more than once a year
• New reagent lot must arrive a month before old lot runs out to complete
lot-to-lot comparison

New lot is rejected when:

• > 10% variance
• All variances are positive
• All variances are negative
 If use of new lot is necessary, new RI & therapeutic range must be developed
Internal Quality Control
• RANGE - the difference between the largest and the smallest values in a
data group
• Interval – a statistics that trim outliers

• Reference interval (RI) = reference range = normal range

• Therapeutic range - concentration usually expected to achieve desired
therapeutic effects.
Internal Quality Control
Developing an RI
• Define population to provide blood specimen
 demographics (age & race)
• Equal number of males & females
• 25 selected subjects (PT & PTT)
• 120 data for new assays with no currently established RI

To Validate manufacturer’s RI
• 15 males & 15 females
 Transference
 Typical RI is computed as mean ±2SD & distribution is normal (Gaussian)
Internal Quality Assurance
 ensure reliable & reproducible results regularly
• daily quality control chart
• written policies & procedures
• work instructions
• calibration reports
• reagent evaluation records
• manpower training
• competency manuals
• safety information
• continuous programs to improve quality
Internal Quality Assurance
Quality Controls
 a system that is set up to ensure that certain limits for a test result or a
product are maintained
 aggregate of processes and techniques to detect, reduce & correct
deficiencies in an analytical process
Involves
• Analysis of control sample
o result should fall within the predetermined range
• Statistical evaluation of results
o to determine the acceptability of the analytical run
Internal Quality Assurance
1. Quality Controls
Automated Methods
• To monitor the performance of the automated cell counters on a continuing basis
• Daily / twice a day/ per shift
 every 100 tests
• 2 levels of control materials
 normal
 pathologic/abnormal
Types of control materials
• Commercial control material
• Fresh whole blood control
Internal Quality Assurance
Types of control materials
1. Commercial control material
• blood cells (fixed, buffered, stabilized or preserved)
• human, avian porcine
• “WBC” – nucleated avian red cells
2. Fresh whole blood control
• ideal for control because it is identical to the material being tested
o kept refrigerated
o used w/in 24 hrs
• Hgb is table for several days
• platelets & leukocytes are affected by aging
Internal Quality Control
Quality Control Monitor

1. Levey-Jennings Chart
• Graphically display the assay values of controls versus time
• Assumes that the control results distribute in a Gaussian
• Indicates the mean, 1,2, and 3 SD on both side of the mean
Internal Quality Control
Levey-Jennings Chart

Systemic drift/trend
• control value moves progressively in one direction from the mean for at
least 3 days
• problem progressively developing
• Deterioration of reagent or control
• Diluent contamination
Internal Quality Control
Levey-Jennings Chart
Dispersion
Causes
• Random errors
• Lack of precision
Indicate
• Inconsistency in technique
• Stability problem
Internal Quality Control
Levey-Jennings Chart

Shift

• Abrupt change

• Associated with
• Malfunction of instrument
• Error in technique
Internal Quality Control
Multirule (Westgard) Analysis
 series of multirules to help evaluate paired control runs, one low and one high
concentration at the end of the test method’s linearity range
• 12S rule – a control value is outside a 2s limit
• 13S rule – one value is outside a 3s limit
• 22S rule
• 2 consecutive values are outside the same 2s limits. This may be within the same
control run involving both levels of control exceeding the same +2 or -2 limit
• 2 consecutive analyses of the same control materials exceeding the same 2s limit
Internal Quality Control - Multirule analysis

12S 13S

22S
Internal Quality Control
Westgard Multirule System Titles
• R4S rule
 2 consecutive values are more that 4s apart, involving both control
materials.
• 1 control is beyond the 2+ limit, and the other is beyond the -2 limit.
• 41S rule
 4 consecutive values have been plotted on the same side of the 1s range.
• may be within or across control materials
Multirule analysis

R4S rule 41S rule

 Westgard Multirule System
ACTIONS FOR SELECTED WESTGARD RULES

RULE What should be done? Can patient result be reported

12S rule Check the machine and method used YES

13S rule Investigate for random error YES
22S Check the control and method used until control result is acceptable
R4S Check the control and method used until control result is acceptable
Investigate for a shift or trend in the
41S until control result is acceptable
analytical process
Internal Quality Assurance
STEPS TO CORRECT OUT-OF-CONTROL ASSAY RUN
1. Reassay

2. Prepare new control and reassay

3. Prepare fresh reagent and reassay

4. Recalibrate instrument
Internal Quality Assurance
Bull’s Algorithm (Moving average of the RBC indices)

• Developed by Dr. Brian Bull (1974)

• method employing patients RBC indices to monitor the stability of automated cell
analyzer
• MCV, MCH & MCHC
• remains constant on average despite patient variations
• RBC count (common denominator)
• WBC & platelet count
Internal Quality Assurance
Bull’s Algorithm
• trim
• Removing small designated percentage of the largest & smallest value
before calculating the mean
• smooth
• Remove “noise” from a data set
• Trimmed 20 specimen mean is plotted on Levey-Jennings chart
• ±3% or 1SD as the action limit
• Track shift & trends using the westguard rule
Internal Quality Assurance
Bull’s Algorithm
Disadvantage:
1. Does not detect within-run errors
2. Less sensitive than commercial controls in detecting system trends and shifts
3. May generate preponderance of outliers in a population with high
percentage of abnormal hematologic results
4. Requires computer to calculate the average
Moving average system do not replace the use of control specimen but provide
additional means to detect shifts and trends.
Internal Quality Assurance
Manual Methods
Quality control on manual counting is primarily a function of the technical
skill and expertise of the Medical Technologists.
• high percentage of error that cannot be avoided

Attributed to:
• Small number of cell counted
• Uncontrollable variation in cell distribution
• Error in graduation of pipette & hemocytometer
• Operator errors in dilution & loading processes
Internal Quality Assurance
DELTA CHECK SYSTEM
• comparison of current analyte result with the result from the most recently previous result
for the same patient
• Detect random error
• ∂ CHECK FAILURE
• 20% deviation
• investigate for an intervention
• Transfusion, surgery
 No explanation
• analytical error; mislabeled specimen; specimen collection error
“An investigation should be undertaken before results are reported.”
External Quality Assurance
External Quality Assessment

• Validate the accuracy of assays

• Proficiency testing system

• Proficiency testing specimen
• Results should match the predetermined targets
• National External Quality Assessment Scheme (NEQAS)
External Quality Assurance
Laboratory Staff Competence
• Integrity & professionalism
• Keys to assay reliability
• Documentation of errors & incident
• Quality improvement & remedial instructions
External Quality Assurance
Laboratory Staff Competence
• Assess & document professional staff skills
• Examination of PBS
o Personnel missing the target result = remedial instruction
• Continuing Education
o Maintains the critical skills of laboratory personnel
o Provides opportunity to learn about new clinical & technical
approaches