Clinical Chemistry (Quality Control)

Practicability

Method is easily repeated

Reliability

Maintain accuracy and precision

Intralab/Interlab QC

Daily monitoring of accuracy and precision

Interlab/External QC

Proficiency testing (Reference lab)

Long-term accuracy

*Difference of > or = 2 = NOT IN AGREEMENT WITH OTHER LAB

QC materials

available for minimum of 1 year

Bovine Control Materials

Preferred (Human: Biohazard)

NOT for immunochem, dye-binding, and bilirubin

Matrix effect

Due to improper product manufacturing

Unpurified analyte

Altered protein

Precision study

First step in method evaluation

Nonlab. personnel

29% of errors (lab results)

SD (Standard Deviation)

Dispersion of values from the mean

CV (Coefficient of Variation)

Index of precision

Relative magnitude of variability (%)

Variance

SD squared

Measure of variability

Inferential statistics

Compare means or SD of 2 groups of data

T-test

Means of 2 groups of data

F-test

SD of 2 groups of data

Cumulative Sum Graph (CUSUM)

V-mask

Earliest indication of systematic errors (trend)

Youden/Twin Plot

Compare results obtained from different lab

Shewhart Levey-Jennings Chart

Graphic representation of the acceptable limits of variation

Trend

Gradual loss of reliability

Cause: Deterioration of reagents (Systematic error)

Shift

Values: one side or either side of the mean

Cause: Improper calibration (Systematic error)

Outliers

Values: far from the main set of values

Highly deviating values

Random or systematic errors

Kurtosis

Degree of flatness or sharpness

Precision

Reproducibility of a value

Random error

Accuracy

Closeness of value to true value

Systematic error

Random error (Imprecision; Indeterminate)

Causes:

-Mislabeling

-Pipetting error

-Improper mixing of sample and reagents

-Voltage/Temperature fluctuation

-Dirty optics

Parameters: SD and CV

Systematic error (Inaccuracy/Determinate)

Causes:

-Improper calibration

-Deterioration of reagents

-Contaminated solution

-Sample instability/unstable reagent blanks

-Diminishing lamp power

-Incorrect sample and reagent volume

Parameter: Mean

Multirule Shewhart procedure

Control rules + Control Chart

Test method

involves at least 40 samples (Westgard)

Reference method

preferably 100 samples (Westgard

Analytical Run

Control and patient specimens assayed, evaluated, and report together

Physiologic Limit

Referred to as absurd value

POCT (Point of Care Testing)

Performed by nonlab personnel

Quality Assurance

Tripod:

Program development

Assessment and monitoring

Quality improvement

Quality Patient Care

Test Request forms, clear instruction for patient preparation, proper specimen handling, etc.

Reference Range/Interval Range/Reference values

At least 120 individuals should be tested in each age and sex category