Expert Says Sample Sizes Need to be Larger to Satisfy FDA Expectations
Manufacturers of drugs and medical devices make several common errors when choosing the sample size for their sampling plan, but selecting too few samples routinely tops the list, according to Steven Walfish, president of Statistical Outsourcing Services.
“The most common error I see is the fact that the sample size is too small for the risks that you’re trying to get at,” he said in a recent webinar hosted by FDAnews, a WCG company.
The samples may show a 99 percent reliability but only 50 percent confidence in the results. “What happens is that we’ve lowered our confidence level” by having too few samples, he said. “We tend to trade confidence and reliability off each other in order to justify our sample size.”
Access the full webinar, Calculating Sample Size to Satisfy FDA Expectations, here.
To read the whole story, click here to subscribe.