Main menu


Automation in Hematology (4)

Automation in Hematology (4)

Quality Control / Quality Assurance/Quality Assessment 

- Quality Control: measures that must be eclosed during each assay run that the test is working properly. 

- Quality Assurance => overall program that ensures that the final results reported by the laboratory are correct. 

- Quality Assessment => means to determine the quality of results generated by the laboratory.

Quality control

- The measure of precision – how well the measurement system reproduces the same results over time and under varying operating conditions. 

- Designed to detect, reduce & correct deficiencies in a laboratory’s internal analytical processes before release of patient’s reports. 

- Should approximate the same matrix as the patient’s specimens, considering properties such as viscosity, turbidity, composition, and color. 

- Should be simple to use, with the minimal vial to vial comparability. 

- Should be stable for long periods and obtainable in large enough quantities for a single batch to last at least 1 year.

Types of Quality Control

Internal Quality Control:

- Continuous evaluation of the reliability of the daily works of the laboratory with the validation of tests. 

- The primary tool required is called a control – a specimen with a predetermined range of result values, processed in the same manner as a patient sample. 

- If the result of a test on a control sample is different from its known value, this means a problem in the equipment or the methods being used. 

External Quality Control

- Evaluation by an outside agency of the comparability of a laboratory's testing to a source outside the laboratory.
- This comparison can be made to the performance of a peer group of laboratories or the performance of a reference laboratory.
- The analysis of performance is retrospective.

Accuracy and Precision

- Accuracy is a measure of rightness.
- It refers to closeness to the true value.
- Precision is a measure of exactness.
- It refers to the reproducibility of the test. 

- A solution that contains the same components as those being analyzed in the patient sample.
- Trading produced pooled RBCs/sera or stabilized anticoagulated whole blood.
- Should have the same test properties as that of a blood sample.
- At minimum one control specimen should be used for every batch.
- If enormous specimens, use one control for every 20 specimens.
- For most tests, “normal” control and an “abnormal” control are analyzed with each patient's test or batch of tests.
- The results are compared with the manufacturer’s range of values and plotted on a Levey-Jennings Chart.
Levey Jennings Graphs 

- A graphical procedure for displaying control results and evaluating whether a procedure is in-control or out-of-control
- Control values are plotted versus time. 
- Lines are drawn from point to point to confirm any trends, shifts, or random excursions.
- Consecutive values of control are recorded and the standard deviation is calculated.
- The mean and ± 2 SD are layed on a Levey- Jennings chart. 

- As long as the control value is between the ± 2 SD lines on the L-J chart, the control values are “in control”.
- If they are out of the ± 2 SD lines, they are 'out of control'.

Westgard Control Rules 

- Proposed by Dr. James Westgard on lab quality control. 

- Put the basis for evaluating analytical run for medical laboratories. 

- There are 6 basic rules. 

12s Rule. Warning rule to provoke careful inspection of the control data. 

- A single control measurement great than two standard deviations of control limits either above or below the mean.
- No cause for rejecting a run. 

13s rule
- A single control measurement great than the mean plus 3s or the mean minus 3s control limit.
- The run must be rejected. 

22s rule
- Two consecutive control measurements greater than the +2SD or -2SD control limit.
- The run must be rejected. 

R4s rule
- One control measurement exceeds the +2SD and the other exceeds the -2SD control limit.
- The run must be rejected. 

41s rule
- 4 consecutive control measurements are greater than the +1SD or -1SD control limit.
- The run must be rejected. 

10x rule
- 10 consecutive quality control products for one level of control are on one side of the mean.
- The run must be rejected. 
Out of Control
- Stop testing.
- Identify and correct the issue.
- Repeat testing on patient samples and controls.
- Don’t announce patient results until the issue is sorted out and the controls indicate proper performance.
- Systematic error is evidenced by a change in the mean of control values.
- The change in mean maybe
▪ Gradual - demonstrated as a trend in control values.
▪ Abrupt – demonstrated as a shift in control values.
   - Random error is any deflection away from the expected result. 


A gradual dropping of reliability in the test system. 

- Deterioration of instrument light source.
- A gradual accumulation of debris in sample/reagent tubing and electrode surfaces.
- Aging of reagents.
- Gradual deterioration of –
▪ Control materials. 

▪ Incubation chamber temperature. 

▪ Light filter integrity. 

▪ Calibration. 


Sudden/dramatic +ve or –ve change in a test system
- Abrupt failure or change in the light source.
- Change in reagent formulation.
- Major instrument maintenance.
- Sudden change in incubation temperature.
- Change in room temperature or humidity.
- Failure in the sampling system. 
- Failure in reagent dispense system.
- Inaccurate calibration/recalibration.
- discover the accuracy and precision of the analyzer using a precisely prepared product to recover each parameter within close tolerances of known target values and limits.
- Calibrate the hematology analyzer to provide the most accurate results possible.
- Coefficients of variation and percent difference recovery should be within their specified ranges.
Why Calibrate?
Calibration is necessary
- To provide readings from an instrument are consistent with other measurements.
- To ascertain the accuracy of the instrument readings.
- To initiate the reliability of the instrument i.e. that it can be trusted.
When to Calibrate?
- At installation.
- After the replacement of any component that involves dilution characteristics of the primary measurements (such as the apertures).
- When advised to do so by your service officer