In microbial enumeration practice it is recognized a general uncertanity of 0,5 log on the plate count readings (CFU). Thus I think that the example refer to the condition which, given a declared potency at the end of the shel-life, any results that are within the 0,5 log uncertainty are conform and not out of spec.
In industrial real life practice, becasue of this issue, the producer overdose the product at t0 both to encompass reduction in stability and method variability.
Hope it helps!
|