Sampling – Medical Devices

Does the US Food and Drug Administration (FDA) ever make a mistake? In the Preamble to the Quality System Regulation (21 CFR 820, Comment 85), they addressed sampling for Process Validation. Comment 85 says:

The requirement for testing from the first three production lots or batches has been deleted?. The number three is, however, currently considered to be the acceptable standard.

The FDA now admits that this guidance on a sample size of three is NOT statistically valid. They wish they could delete Comment 85 from the Preamble; however, to do this would take an act of the US Congress!

This training is targeted to professionals concerned with a specific application of valid statistical techniques to satisfy Process Validation requirements of the FDA:s Quality System Regulation. Sampling a product or process and measuring a quality characteristic is an efficient way to make an inference about the quality of all products or the entire process. Simply put, if it is too expensive to measure everything, then you would be wise to sample and measure a subset. If you decide to sample, then comes the age-old question: How much is enough?

This fast-paced training will cover the details of application without the burden of statistical complexity. Analysis results from popular statistical software programs will be illustrated. The presenter´s keep it simple approach will avoid statistical complexity – participants do not need to have a background in statistics.

The key to establishing Sample Size is to quantify the risks involved in making a decision on the quality of a product or process. At the end of this webinar, participants will be able to:

• Create a risk-based Sampling Plan for a new production process
• Apply fundamental statistical methods and tools for analyzing a production process
• Describe the factors that influence sample size selection
• Formulate risk-based decisions on the quality of a product or process
• Express the risk involved in a data-based decision

There are no prerequisites for this training. No background in statistics is necessary.

Content

• Regulatory Context
• Quality System Regulation 21 CFR 820 and Preamble Comments
• GHTF Guidance for Process Validation
• Significant Trends and Warning Letters
• Demonstration of Best Practices
• Estimating Process Capability
• Graphical Tools
• Common Pitfalls
• Designing a Sampling Plan for New Production Processes
• A Checklist Approach
• Applying Statistical Methods and Tools
• Sample Size Considerations
• Applying Valid Statistical Software
• Expressing a Risk-Based Decision

Duration: 2 hours

#8009