Hi Bev, I take your point, but I don't think there's any confusion.
(except for the confusion caused by inconsistent expectations/policies from regulators
)
The original poster was clearly expressing the FDA was concerned with their design verification sample sizes and was seeking advice with respect to "electrical safety tests, software verification, mechanical durablility and loadability tests"...most of which are covered in standards like IEC 60601.
So my question stands: Why does the FDA accept IEC sample size of 1, and yet requires justification for internal design verification sample sizes? Does IEC have a "statistically valid" rationale for only requiring one representative sample?
It seems to me that if the FDA will recognize an IEC sample size of 1, then any comparable internal verification activities should also be able to use 1. If more than 1 is used, great...but this shouldn't have to be justified statistically.
(except for the confusion caused by inconsistent expectations/policies from regulators
The original poster was clearly expressing the FDA was concerned with their design verification sample sizes and was seeking advice with respect to "electrical safety tests, software verification, mechanical durablility and loadability tests"...most of which are covered in standards like IEC 60601.
So my question stands: Why does the FDA accept IEC sample size of 1, and yet requires justification for internal design verification sample sizes? Does IEC have a "statistically valid" rationale for only requiring one representative sample?
It seems to me that if the FDA will recognize an IEC sample size of 1, then any comparable internal verification activities should also be able to use 1. If more than 1 is used, great...but this shouldn't have to be justified statistically.
...Sorry, but my cynical side has to chuckle at this...