Cleaning and Disinfection Validation Sample Size of 3?

P

Phloogy

I will be outsourcing a cleaning validation study as well as a disinfection validation study for our medical device. AAMI TIR12 and TIR30 both say to use at least three replicates and that three is relatively standard.

From TIR30 7.3.2 -
NOTE—A minimum of three replicates and one concurrent positive control and one concurrent negative control is relatively standard in validation studies.

Why is three called out when it isn’t statistically significant? Wouldn’t we need a lot more considering this is attribute data?
 

Ronen E

Problem Solver
Moderator
I think that by "replicates" they mean test groups, not test units, but I don't have these documents in front of me.

It is common to do "3 repetitions" in validation studies, meaning repeating something 3 times (3 batches etc.).
 
P

Phloogy

I think that by "replicates" they mean test groups, not test units, but I don't have these documents in front of me.

It is common to do "3 repetitions" in validation studies, meaning repeating something 3 times (3 batches etc.).
Ronen,
I went back to those standards to verify if "replicate" was defined or what was meant by it. In this case, it appears the word replicate is synonymous to test sample. The standard uses the word sample to indicate what was extracted (e.g. swab sample).

I got a response back from a test lab last night and they have not had any issues with the FDA regarding studies with samples ranging from 3 to 11 while not being statistically significant. I just have difficulty providing a statistical rationale for such a small sample size, unless I don't provide one altogether and cite the standard instead. Thoughts?
 
Top Bottom