Impurity Specs

[jayroseville58]

Is it really always needed that when your impurity specs is below 1.0% the criterion should be in two decimal places? Why is it the USP don't follow the same suit? Thanks.

[tom jupille]

Not always, but, imagine for a moment that the spec is "less than 1.0%". If the actual amount is, say, 0.96%, and your computation is carried out to one decimal place, it would round to 1.0% and be a "fail" (remember, the spec is less than 1%). If your computation is carried out to two decimal places, it would be a "pass".

[Kreall]

That's why in Ph. Eur or USP the limit for impurities is "not more than 1,0%", not "less than 1,0%". So if the amount of impuritiy is 0,96%, after rounding up, it is still in the given limit.

So according to Ph. Eur and USP if there is limit with one decimal place, you should give you result with one decimal place. It is only a GLP to give result with more decimal places.