Lean Six Sigma Notes

Lean vs Six Sigma

• Lean Principles:
  • Focus on eliminating waste, inspired by Toyota's practices.
  • Critical Questions:
  • Does this step need to exist?
• Six Sigma:
  • Aims to improve process performance by eliminating the causes of defects and errors, as demonstrated by Motorola.
  • Critical Question:
  • How can this process be improved?

Understanding Waste

• Consequences of Waste:
  • Increased costs
  • Delays in processes
  • Limited efficiency
• Lean Six Sigma Strategy:
  • Employs a problem-cause-solution approach to improve processes.
  • Methodology: DMAIC (Define, Measure, Analyze, Improve, Control)

Sigma Metrics

• Sigma Definition:
  • Quantitative measurement of error or variation in a system.
  • Process sigma indicates the capability of a process to meet or exceed defined acceptance criteria.
  • Key Metrics:
  • Number of defects per million opportunities
  • Number of Standard Deviations (SD) from mean a process can shift before exceeding acceptable limits.
• Example:
  • If a sodium test has a six sigma quality, the mean could shift by 6 SD and still meet accuracy and precision requirements.

Sigma and Errors

• 6 Sigma Quality:
  • Very precise; produces 3 errors for every million tests reported.
• 3 Sigma Quality:
  • Produces 26,674 errors per million tests.

Importance of Sigma Beyond the Analytical Stage

• 12% of lab errors impact patient health.
• Only 0.0375% of errors occur in the analytical stage.

Sigma and Quality Control (QC)

• Excellent precision and accuracy lead to fewer errors.
• A significant mean shift is necessary to fail quality requirements.
• Fewer QC rules may be needed to identify errors.
  • False Rejection: Maximizing the chance of detecting a problem while reducing the risk of incorrectly discarding a correct result.
• Sigma Representation: \text{Sigma} = \mu - \text{(mean)}

Goal of Six Sigma for World-Class Quality

• Tolerance Specification:
  • Aiming for 6 SDs within specification limits.
• Target Specification:
  • Ensures +6 SDs fit into specification.
  • Visual representation of distribution enables understanding acceptable limits.

Minimum Acceptable Performance Levels

• 3 Sigma:
  • Represents minimum performance standards with an increase in the number of allowable defects.

Accuracy and Precision Metrics

• Assess the observed %Bias and %Coefficient of Variation (CV)
• Create a decision chart to classify performance levels based on Sigma levels (2, 3, 4, 5, 6).

Example Sigma-Metric Calculation

• Cholesterol Study Example (Clin Chem July 2014):
  • Acceptability criterion: 10% (CLIA PT criterion).
  • Calculating Total Precision (CV):
  • Bias values: 1.0%, 0.9%, 1.0%
• Sigma Calculation:
  • For 1.0% Bias: \text{Sigma} = \frac{(10 - 3)}{1.0} = \frac{7.0}{1.0} = 7.0
  • For 0.9% Bias: \text{Sigma} = \frac{(10 - 2.5)}{0.9} = \frac{7.5}{0.9} = 8.3
  • For 1.0% Bias: \text{Sigma} = \frac{(10 - 2.3)}{1.0} = \frac{7.7}{1.0} = 7.7
• Average Sigma:
  • \text{Average Sigma} = \frac{(7.0 + 8.3 + 7.7)}{3} = 7.67

Westgard Sigma Rules

• Data Quality Control (QC) guidelines including:
  • Conditions for reporting results
  • Recommendations for corrective actions based on Sigma metrics.
• Sigma Scale Formula: \text{Sigma Scale} = \frac{(%TEa - %Bias)}{%CV}
  • Utilize the Sigma Scale to locate and understand the performance standards.