CLINICAL CHEMISTRY

LECTURE / FIRST SEMESTER

UNIT 2: QUALITY CONTROL
3. Most commercially prepared QC materials are
Topic Outline:
lyophilized (powdered / pulverized form) so that they
 Quality Assurance and Quality Control
require reconstitution before used. Reconstitution is
 Types of Error
the addition of diluent like distilled water.
 Basic Statistics
4. QC for general chemistry assays uses 2 levels of
 Types of Chart
control
 Westgard Rules
 Level 1 or Normal Control - contains normal
levels for the analyte being tested
Quality Assurance and Quality Control  Level 2 or Abnormal Control - contains levels
Quality Assurance (QA) below or above the concentration of analyte of
- is a complete system of creating and following interest
procedures and policies: A. Immunoassays commonly use 3 levels of control
 to aim for providing the most reliable patient laboratory  Level 1 or Low Control - contains levels
results, and below the normal range of the analyte of interest
 to minimize errors in the pre-analytical, analytical, and  Level 2 or Normal Control - control normal
postanalytical phases levels for the analyte being tested
 Level 3 or High Control - contains level
Quality Control (QC) higher than the upper limit of analyte of interest
- a system of ensuring accuracy and precision in the
laboratory by using quality control materials in every series of Example of a QC Log with Patient Results
measurement

Quality Management (QM)

- refers to the overall process used to ensure that
laboratory results meet the requirements for health care services
to patients.

Objectives of Quality Control

1. To check the stability of the machine
2. To check the quality of the reagents
3. To check technical (operator) errors

Operation of a Quality Control System

1. Establishing allowable statistical limits of variation (error) for
each analytic method
2. Using these limits as criteria for evaluating the QC data
generated for each test
3. Taking action to remedy errors when indicated
a) Finding the cause(s) of error
b) Taking corrective action
c) Reanalyzing control and patient data

Kinds of Quality Control

Quality Control 1. Intralab Quality Control (Internal QC)
Control Limits - involves the analyses of control samples and the
- are intervals of acceptable values with upper and patient specimen; use for the daily monitoring of accuracy and
lower limits. precision of analytical methods
1. Values fall within the control limits – the analytic
method is properly reporting values 2. Interlab Quality Control (External QC)
2. Values fall outside the control values – identify - involves proficiency testing programs that periodically
possible problems and do corrective measures provide samples of unknown concentrations to participating
3. Ideal control limit - ± 2SD clinical laboratories; important in maintaining long-term accuracy
- The principles of analyzing QC in the laboratory were of analytical methods
applied in 1950s by Levey and Jennings. - to determine state-of-the-art interlaboratory
performance
QC Materials (Control Solutions) - College of American Pathologists (CAP) proficiency
- are specimens analyzed for QC purposes. program - gold standard for clinical laboratory external QC
1. Available in sufficient quantity to last at least a year testing
2. The same matrix as the specimens to be tested - NEQAS (National External Quality Assurance
Scheme)
 CLINICAL CHEMISTRY
LECTURE / FIRST SEMESTER
NEQAS (National External Quality Assurance Scheme) Diagnostic Sensitivity
- As provided in Department Order No. 393-E s. 2000, - ability of a test to detect a given disease or condition
five institutions were designated as the National Reference TP
Laboratories namely:
1. Research Institute for Tropical Medicine (RITM) TP+ FN
- NRL for dengue, Influenza, TB, and other
Mycobacteria, Malaria and other parasites, Bacterial enteric Diagnostic Specificity
disease. Measles and other viral exanthems, Mycology, - ability of a test to correctly identify the absence of a
Enteroviruses, Antimicrobial resistance and engineering given disease or condition
diseases TN
TN + FP
2. San Lazaro Hospital / STD AIDS Cooperative Central
Laboratory (SLH/SACCL)
False Negative
- NRL for HIV/AIDS, Hepatitis, Syphillis and orher STI's,
- a patient has the disease but is negative of the test
Immunology and Serology
False Positive
- the patient does not have the disease but is positive of
3. National Kidney and Transplant Institute (NKTI)
the test
- NRL for Hematology inc. Immunohematology and
human chorionic gonadotropin (HCG)
Immunopathology and Anatomic Pathology
TP = True Positive FP = False Positive
4. Lung Center of the Philippines (LCP)
TN = True Negative FN = False Negative
- NRL in in atomic pathology for pulmonary and pleural
disease, Clinical Chemistry

5. East Avenue Medical Center (EAMC)

- NRL for environmental and occupational health;
toxicology and micronutrient assay

Ten Analytes Measured for NEQAS in Clinical Chemistry

1. Glucose
2. Blood Urea Nitrogen (BUN)
3. Creatinine
4. Uric Acid
5. Cholesterol
6. Triglyceride
7. Albumin Types of Error
8. Sodium 1. Random Error
9. Potassium - present in all measurements; due to chance
10. Chloride - due to instrument, operator and environmental
conditions such as pipetting error, mislabeling in samples,
Parameters of Quality Control temperature of analyzer, and improper mixing of sample or
1. Sensitivity reagent
- ability of the method to measure the smallest
concentration of the analyte of interest, limit of detection 2. Systematic Error
2. Specificity - error always in one direction
- measure only the analyte of interest - calibration problems, deterioration of reagents and
3. Accuracy control materials, improperly made standard solutions,
- nearness or closeness of the assayed value or the contaminated solutions, unstable and inadequate reagent
true value or target value blanks, leaky ion selective electrode, poorly written procedures
4. Precision  Constant Error
- the closeness of the assayed value to the repeated - the magnitude of change is constant and not
value dependent on amount of analyte
5. Practicability - cause: interference or contamination
- degree by which the method is easily repeated  Proportional Error (Slope/Recent Error)
6. Reliability - error dependent on analyte concentration
- ability of the analytical method to maintain accuracy - cause: poor recovery of analyte during an analysis
and precision over an extended period of time
3. Allowable Analytic Error (Ea)
Diagnostic Sensitivity and Specificity
Diagnostic Efficiency - If RE and SE (total error) < Ea, then the performance
- used to determine how good the given test is at of the test is considered acceptable.
detecting the presence of a disease - If the error > Ea, corrections must be made to reduce
the error or the method is rejected.
- Published by the Clinical Laboratory Improvement
Amendments of 1988 (CLIA 88)
 CLINICAL CHEMISTRY
LECTURE / FIRST SEMESTER
values will be squared thus eliminating the significance of
the plus or minus values.
4. Square the difference of each

5. Total the differences squared

6. Divide this sum by one less than the total no. of observed
values and take the square root.

Example:

Basic Statistics
1. Mean
- a measure of central tendency, average

Mean=
∑x
n
Where x = observed value; n= number of observed values

2. Standard Deviation
- the most frequently used measure of variation
- a measure of dispersion of values from the mean;
helps describe the normal curve
- Low SD: the values tend to be very close to the mean
- High SD: the values are spread out, greater the
variation, more error

3. Coefficient of Variation (CV)

- an index of precision
- the higher the CV, the greater is the level of dispersion
around the mean
- the lower the CV, the more precise is the estimate
- CV of high precise analyzers can be lower then 1%

4. Variance (V)
- measure of variability
2
V =( SD)
Procedure for Calculation of SD:
1. Perform at least 30 analyses on the control sera; these
should be done daily for each determination until 30
observed values are obtained. Record all values.
2. When 30 values are obtained, add up all the values and
calculate the arithmetic mean.

3. Take the difference between the mean and each of the

observed values. Disregard negative or positive values, the
 CLINICAL CHEMISTRY
LECTURE / FIRST SEMESTER

2. Shewhart Levey-Jennings Chart

- “S-L/J chart, dot chart”
- It is the most widely used QC chart in the clinical
laboratory
- Can apply multiple rules without the aid of the
computer
- Easily identifies random and systematic errors
Other Types: Cumulative Sum Graph; Youden / Twin Plot

Errors which can be observed on the LJ Chart:

A. Trend
- formed by the control values that continue either to
increase or decrease for a period of six consecutive days by
passing the mean
- Main cause: deterioration of reagents

Quality Control Chart B. Shift

Types: - Formed by the control values that distribute
1. Gaussian Curve themselves on one side of the mean for a period of six
- (Bell-shaped curve / Normal distribution curve / consecutive days
Normal frequency curve) - Main cause: improper calibration of the instrument
- Ideal control limit for clinical chemistry is ± 2 SD
C. Outliers
- Are values that are far from the main set of values
- Are highly deviating values
 CLINICAL CHEMISTRY
LECTURE / FIRST SEMESTER
- Caused by R/S errors - This allows the detection of systematic error

Westgard Rules E. R4s

In 1981, Dr. James Westgard of the University of - The range or difference between the highest and
Wisconsin published the six basic Westgard Rules and lowest control result within an analytical run exceeds +2SD and
these rules are used to further judge whether the analytic -2SD
run or measurement is out of control or in control. - Allows detection of random error

1 - Control Observation; 2s - statistical limit in subscript

A. 12s
- One control observation exceeds the mean + 2SD.
- A warning rule that initiates testing of control data by
other rules. F. 10x
- for screening purposes - Ten consecutive control observations are on the same
side of the target mean.
 - This allows the detection of systematic error

B. 13s
- One control observation exceeds the mean + 3SD
- Allows high sensitivity to random error

C. 22s
- Two control observations consecutively exceed either
the mean ±2SD
- Allows high sensitivity to systematic error

D. 41s
- Four consecutive control observations exceed
either mean ±1SD 70% of all medical decisions are based on laboratory results
(Silverstein, 2003)
 CLINICAL CHEMISTRY
LECTURE / FIRST SEMESTER

Quality is everyone’s responsibility.