LESSON 5: QUALITY ASSURANCE-QUALITY CONTROL

What is Quality Assurance?

All processes a laboratory takes to ensure quality in test results. It's a quality management process that establishes standards, guidelines, and procedures to prevent quality issues.

What percentage of errors occur in the pre-analytical phase?

70% of laboratory errors occur in the pre-analytical phase, which involves patient interaction, specimen collection, and transport.

What are the three phases of Quality Assurance?

1. Pre-Analytical (patient interaction, specimen collection, transport) 2. Analytical (test performance, reagent handling, instrument maintenance) 3. Post-Analytical (result verification, reporting, physician interpretation)

What are the 4 steps of the Quality Management Framework?

1. Quality Planning 2. Quality Assurance 3. Quality Control 4. Quality Improvement

What is the difference between Quality Control and Quality Assurance?

Quality Control is limited to the analytical process only and uses QC reagents to detect problems early. Quality Assurance encompasses all laboratory processes from pre-analytical to post-analytical phases.

What is the difference between Random Error and Systematic Error?

Random Error: Non-recurring, unpredictable errors (usually medtech-related). Systematic Error: Recurring, predictable errors that affect all results equally (usually analyzer-related).

What are common causes of Random Error?

Mislabeling, pipetting errors, improper mixing, voltage/temperature fluctuations, dirty optics, presence of bubbles in reagent/sample lines.

What are common causes of Systematic Error?

Improper calibration, deterioration of reagents/calibrators, contaminated solutions, sample instability, diminishing lamp power, incorrect sample volumes, wrong calibrator values.

What steps should be taken when control values fall outside acceptable range?

1. Don't release patient results 2. Repeat the control 3. Run newly reconstituted control 4. Look for obvious problems 5. Recalibrate instrument 6. Install new reagents if needed 7. Perform maintenance 8. Document everything

What is Internal QC vs External QC?

Internal QC: Daily running of QC material within lab, statistical monitoring using Levey-Jennings charts. External QC: Third-party proficiency testing (NEQAS, PCQACL) comparing labs nationally.

What are the characteristics of good QC materials?

Stable, homogenous, available in aliquots. Should have different levels (L1: low-normal, L2: normal-high, L3: significantly elevated) to test analyzer capability across ranges.

What is the difference between Accuracy and Precision?

Accuracy: Ability to obtain the established 'true' value for a sample (relates to systematic error). Precision: Ability to obtain the same value for repeat measurements (relates to random error).

When should Quality Control be performed?

CLIA recommends QC at least once daily for Level 1 and Level 2 controls. If only Level 1 available, perform twice daily (morning and night shift). Often done during night shift due to lower patient workload.

What is a Delta Check?

Assessment comparing a patient's current test result to their previous value for the same test. The difference (delta) is calculated and compared to established limits to detect errors.

What are the steps to implement QC?

1. Obtain control material 2. Run each control at least 20 times over 30 days 3. Calculate mean and SD 4. Plot values on Levey-Jennings Chart

What is the goal of Quality Control?

To ensure precision and accuracy by using QC reagents in each series of measurements. Monitor machine stability, reagent quality, and operator performance to prevent errors.

What is the 1₂S rule?

WARNING RULE: When one control value exceeds the mean by more than 2SD but less than 3SD. This is a warning to proceed with caution, not a rejection rule.

What is the 1₃S rule?

REJECT RULE: When one control measurement exceeds the mean by ±3SD. Usually indicates random error caused by pipetting mistakes, air bubbles, or technical issues.

What is the 2₂S rule?

REJECT RULE: When two consecutive control measurements exceed the same mean ±2SD. Usually indicates systematic error and requires stopping testing and investigating the source.

What is the R₄S rule?

REJECT RULE: When the difference between two consecutive controls is greater than 4SD. Usually indicates random error with results scattered in opposite directions.

What is the 4₁S rule?

REJECT RULE: When four consecutive control measurements exceed the same mean ±1SD. Usually indicates systematic error caused by instrument malfunction or reagent deterioration.

What is the 10x rule?

REJECT RULE: When ten consecutive control measurements fall on one side of the mean. Indicates systematic error with persistent bias in the testing system.

What is the Mean in statistics?

The average of all data points. Formula: Sum of all data values ÷ Number of data values. It relates to accuracy and systematic error.

What is Standard Deviation (SD)?

A measure of variation that describes the average distance each data point is from the mean. Smaller SD = better precision, closer to mean. Larger SD = wider distribution, greater random error.

What is Coefficient of Variation (CV)?

Standard deviation expressed as a percentage of the mean. Formula: (SD ÷ Mean) × 100. Acceptable CV should be less than 5%.

What percentage of values fall within ±1SD, ±2SD, and ±3SD?

±1SD = 68.2% of values ±2SD = 95.5% of values ±3SD = 99.7% of values (Normal Gaussian Distribution)

What is a Levey-Jennings Chart?

A quality control chart that plots QC values over time. X-axis shows days/runs, Y-axis shows QC values. Mean is center line with ±1SD, ±2SD, ±3SD lines marked.

What are the warning and rejection limits on a Levey-Jennings chart?

Warning limits: ±2SD (analysis should be repeated) Rejection limits: ±3SD (analysis should stop, patient results held, system investigated)

What is a TREND in QC?

A systematic drift in one direction away from the mean for 6 consecutive runs. It's a gradual change indicating loss of reliability, often caused by reagent deterioration.

What is a SHIFT in QC?

A sudden switch of data points to another area away from the previous mean. It's an abrupt, continuous change often caused by improper calibration.

What is DISPERSION in QC?

When control values are widely scattered in an unexplained pattern around the control chart. Can indicate malfunctioning analyzer parts like automatic dispensers.

What is Sensitivity vs Specificity?

Sensitivity: Percentage of diseased individuals correctly identified as positive (TP÷(TP+FN)). Specificity: Percentage of non-diseased individuals correctly identified as negative (TN÷(TN+FP)).

What is a Youden Plot?

A graphical technique for analyzing interlaboratory data when each lab has made two runs on the same analyte. Identifies within-laboratory and between-laboratory variability.

What are Confidence Intervals in QC?

Limits between which we expect a specified proportion of population to lie based on normal distribution: ±1SD = 68.2%, ±2SD = 95.5%, ±3SD = 99.7%

What is Range in statistical analysis?

The difference between the highest and lowest value in a data set. Formula: Largest value - Smallest value. Arrange data in order to find these values.

What is Six Sigma methodology?

A manufacturing strategy pioneered by Motorola in the 1980s to decrease defect rates. Goal: Only 3.4 defects per million opportunities (99.99966% success rate).

What does DMAIC stand for in Six Sigma?

Define (identify the problem) Measure (document baseline data) Analyze (identify root cause) Improve (implement new processes) Control (ensure new processes are maintained)

What are the Six Sigma belt levels?

White Belt (basic intro) Yellow Belt (team member) Green Belt (team leader) Black Belt (expert) Master Black Belt (trains black belts) Champion (high-level executive)

How do you calculate DPMO?

DPMO = (Number of defects ÷ (Sample size × Number of defect opportunities)) × 1,000,000 Used to convert problems into mathematical equations for Six Sigma analysis.

What is the Sigma Metric formula?

σ metric = (TEa - Bias) ÷ CV Where TEa = Total allowable error, Bias = difference between lab means, CV = coefficient of variation

What is Total Allowable Error (TEa)?

Error within acceptable limits as determined by CLIA '88 proficiency testing criteria. Different analytes have different TEa values (e.g., creatinine ±15%, cholesterol ±10%).

What is BIAS in Six Sigma calculations?

The difference between group mean and lab mean, expressed as percentage: Bias = ((Group mean - Lab mean) ÷ Group mean) × 100

What is the Method Decision Chart used for?

Used with Sigma Metric calculations to plot CV (x-axis) vs Bias (y-axis). Different colored regions indicate sigma levels from 1σ (unacceptable) to 6σ (world class)

What QC frequency is recommended for different Sigma levels?

6σ: 1 per 1000 samples 5σ: 1 per 450 samples 4σ: 1 per 200 samples 3σ: 1 per 45 samples 2σ and below: 1 per 10 samples

What is Lean Six Sigma?

Combines Lean Principles (eliminate waste) with Six Sigma (remove variation). Asks 'How can this be improved?' and 'Does this process need to exist?' to streamline operations.

What is DMADV in Six Sigma?

Define, Measure, Analyze, Design, Verify. Used for creating new products/processes, while DMAIC is for improving existing processes.

What companies successfully use Six Sigma?

Motorola (earned $17M more), Toyota (quality manufacturing), Boeing (aerospace safety). Critical in industries where defects can be life-threatening or extremely costly

What is the Sigma Conversion Table used for?

Compares calculated DPMO values to determine sigma level and process yield percentage. Shows relationship between sigma level, defects, and success rates.

What are the 2 ways to determine Sigma Level?

1. Sigma Metric: σ metric = (TEa - Bias) ÷ CV, then plot on Method Decision Chart 2. DPMO: Calculate defects per million opportunities and compare to Sigma Conversion Table

What is the goal of Six Sigma in clinical laboratories?

To achieve near-zero defect rates in healthcare, where only 3.4 results out of 1 million would be erroneous, improving patient safety and laboratory efficiency.

What sigma levels correspond to different success rates?

1σ = 31% yield, 2σ = 69% yield, 3σ = 93.3% yield, 4σ = 99.38% yield, 5σ = 99.977% yield, 6σ = 99.99966% yield

How do you plot a Method Decision Chart?

Scale Y-axis from 0 to TEa (bias), X-axis from 0 to TEa/2 (CV). Draw lines at TEa/2=2σ, TEa/3=3σ, TEa/4=4σ, TEa/5=5σ, TEa/6=6σ. Plot CV vs Bias to determine sigma level.

What is Variance?

The average of the squared distances of all values from the mean. Measures dispersion and variability in the data set.

What is the Mode?

The value that occurs with the greatest frequency in a data set. To find it, arrange data in descending order and identify the most frequently occurring value.

What is the Median?

The middle data point when values are arranged in ascending or descending order. If even number of values, average the two middle positions.