QC, Calibration, and Proficiency Testing

What is a calibrator/standard?

a substance with a known concentration or quantity used to calibrate instruments and validate their accuracy

What does calibrating an instrument in the lab do?

• it essentially teaches your instrument to read

• also establishes a linear ranges for assays

When is calibration performed?

• new analyzers

• using new lot number of reagents

• when recommended by the manufacturers

• after QC failures or shifts/trends

What is beer’s law?

a relationship that demonstrates that the concentration of a substance is directly proportional to the amount of light absorbed

What is the analytical measurement range (AMR)?

• range of results in which the instrument response and the corresponding value is linear

• results outside this range cannot be reported

What is the clinical reportable range (CRR)?

• aka extended AMR

• extended reportable ranges that include values obtained by sample dilution

List 4 requirements of QC materials

• must be similar to the patient specimens being tested

• should have long stability (shelf life)

• should have little variation between vials of the same lot number

• must span normal and abnormal ranges

What is liquid pooled QC material?

pooled patient samples used for QC

What are assayed QC materials?

• QC materials that the manufacturer has tested many times to provide a target value and range for the material

• saves time and resources but is expensive

What are unassayed QC materials?

• QC materials that come with no target values or ranges

• individual labs must test their own targets and ranges

• time and resource extensive

• QC is run several times a day and results over 30 days are collected and ranges are estimated

What is the purpose of a mean value in creating QC targets and interpreting QC data?

the mean acts as a target value for the QC materials

the mean is also the point that is most frequently reported

What is standard deviation?

• how widely spread the QC values fall around the mean

  • high SD = wide variation around the mean

  • low SD = small variation around the mean

coefficient of variation formula

CV = SD / mean x 100

• expressed as a percentage

Generally, QC results in the ___ range are acceptable.

+- 2 SD

Levey Jennings chart

• chart that displays how QC values compare to the mean/target value

• used to notice and correct a test system error before it affects patient results

What are QC trends?

• gradual increases or decreases in QC points

• can be caused by reagent deterioration or light source deterioration

What are QC shifts?

• abrupt increase or decrease in QC points

• usually involves switching from one side of the mean to the other

• caused by recalibration, loading a new reagent lot number, or analyzer issues

Causes of random QC errors (7)

• operator technique

• incorrect reconstitution of QC

• not using CLRW

• pipetting mistakes (by the tech)

• power supply issues

• analyzer pipette issues

  • air bubbles (picked up by the instrument’s pipettor)

What are systemic QC errors, and what causes them to happen?

• more significant errors, usually related to the test system or policies and procedures

• causes

  • improperly aligned pipettes in analyzer

  • unstable temperatures in the analyzer

  • changing the reagent lot number or calibrator lot number

  • deterioration of reagents while in use

1_2s Westgard violation

• 1 QC data point is 2 SDs away from the mean

• considered a warning — may indicate random or systemic error

1_3s Westgard violation

• one QC value 3 SDs away

• can be due to random error

• QC MUST be resolved before continuing patient testing

2_s2 Westgard violation

• 2 QC values 2 SDs away

• QC failure, must be resolved

4_1s Westgard violation

• 4 QC data points 1 SD away

• QC failure, must be resolved

10_x Westgard violation

• 10 QC data points are outside the mean (usually on one side)

• QC FAILURE

R4s Westgard violation

• 1 QC data point is +2 SD away and 1 point is -2 SD away in a single run

• QC FAILURE

What happens with QC data every month?

• the data is submitted to the manufacturer

• the manufacturer compiles all results and publishes “peer group” data

  • a collection of labs performing the same test with the same analyzer using the same QC materials and lot numbers

• lab-specific data is compared to peer group data