Selecting & Justifying Sample Sizes: Where to Begin?

The International Organization of Standardization (150), American Society for Testing and Mate­rials (ASTM), and other standards committees are known to provide valuable information regarding how to perform specific tests. However, the sample size requirements to perform these tests are not always included. In most cases, sample sizes have been left to each manu­facturer to select and justify, based on a statistically valid rationale (150 11607-1 section 4.3).

Risk assessments, historical data, internal procedures, confidence intervals and reliability levels, and some industry standards are examples of processes used for selecting the sample sizes for testing. By selecting a sample size at the smaller end of the range, it is possible that regulatory bodies will determine the testing outcomes are insufficient and either require additional testing be performed, or could ultimately reject the submission.

However, selecting a sample size on the larger end may result in unnecessary costs and resources allocated to produce, manage, and test the larger number of samples. Within our industry, we have seen several approaches to selecting samples sizes. This white paper will not cover all of the approaches, but it is a guide to give medical device manufacturers and pharmaceutical companies a good starting point for determining their own ideal sample size.

---