Understanding PT Statistical Analysis and Evaluation

Oral Presentation

Prepared by C. Huff
ERA, 16341 Table Mountain Parkway, Golden, CO, 80403, United States


Contact Information: [email protected]; 303-463-3536


ABSTRACT

Proficiency Testing (PT) is a critical component of the laboratory accreditation process but it can also be utilized as a tool to demonstrate, monitor and improve the quality of data laboratories produce. Understanding how PT data is analyzed and evaluated plays a critical role in the success of any laboratory.
This presentation provides a comprehensive overview of the PT data evaluation process. Key topics discussed are:

- Statistical analysis of the data population (TNI criteria, z- scoring)
- Robust vs. Arithmetic mean and standard deviation.
- Outlier determination and treatment.
- Regression equations vs. fixed limits: Where do they come from and how are they derived?
- Data modality: What is it and how are multi-modal data handled?
- Assigned values: How are they determined?
- Data evaluation: What drives it and what does it mean?
- How are acceptance limits derived?
- Concentration ranges and their role in determining acceptance limits.
- Monitoring and trending your historical PT performance: Tools and value provided.

This presentation is designed to provide the PT participant with answers to many frequently asked questions and provide fundamental knowledge of PT statistics and data evaluation.