58.648 - Microbiological requirements for ice cream - Insight into 3 out of 5
Can someone explain this regulation and the ending statement of "in three out of five samples"? Seems to me the regulation is a bit outdated and doesn't give much assurance when it is only coliforms and then you can "cherry pick" your results? Is there a guidance document associated with this reg that may give additional insight. Does it literally mean that if you get 3 acceptable results of <10 or <20 coliforms with 2 that could be any result you can release? This is a bit mind boggling, but maybe I am missing some context. Please educate me a bit if you have insight.
It's pity I'm not familiar with FDA regulations RE: ice cream, only Canadian... One of my companies is ice cream manufacturer, and I've been with them for 9 years. AFAK, standardized ice cream products in Canada fall under requirements of having Coliforms not more than 10 cfu/g without specifying quantity of samples tested
Can someone explain this regulation and the ending statement of "in three out of five samples"? Seems to me the regulation is a bit outdated and doesn't give much assurance when it is only coliforms and then you can "cherry pick" your results? Is there a guidance document associated with this reg that may give additional insight. Does it literally mean that if you get 3 acceptable results of <10 or <20 coliforms with 2 that could be any result you can release? This is a bit mind boggling, but maybe I am missing some context. Please educate me a bit if you have insight.
the standard is -
The finished product shall contain not more than 50,000 bacteria per gram as determined by the standard plate count, and shall contain not more than 10 coliform organisms per gram for plain and not more than 20 coliform per gram in chocolate, fruit, nut or other flavors in three out of five samples
Assuming integer values, it "could" mean that for "plain" the worst possible, acceptable, result would be 9.9,10,10,10
It's vaguely equivalent to a conventional nmMc format with n = 5, M=10, m = (1-9) and c=3
The plate count is more fuzzy though, ie average ?
Regardless, I agree with the OP that the "Rule" is definitely in need of an overhaul.
This goes back to FDA standard of identity for Ice Cream and how it is regulated / enforced under the Pasteurized Milk Ordinance.
https://www.accessda....cfm?fr=135.110
https://www.fda.gov/...114169/download
The 3 of 5 samples refers to the regulatory officer, inspector, and the samples they take and send to the state / county laboratory for analysis. If they find more than 3 of 5 consecutive samples they take to be outside the limits they can pull your permit. Typically, this does not happen, rather you get another friendly visit from the inspector and have to deal with a notice of violation where you have to apply a corrective action.
The inspector also gets dinged for it when they are check rated by another inspector where they evaluate everything from your facility, and also, everything the inspector has completed or done for your facility.
I am assuming you are a Grade A facility and have your inspector coming in at least quarterly? Are they taking samples of your products?