Coliform limit in USA dairy products (ice cream)

Got an ice cream product which I made a 10x and 100x dilution and inoculated 1ml of each on coliform plate and after incubation I got 4 and 1 counts respectively. Meaning the cfu is 40cfu/g and 100cfu/g respectively. Which is above the recommended limit of 10cfu/g. Is the a different method of doing dilution because it seems difficult get the recommended limit after several analysis on different ice cream

Got an ice cream product which I made a 10x and 100x dilution and inoculated 1ml of each on coliform plate and after incubation I got 4 and 1 counts respectively. Meaning the cfu is 40cfu/g and 100cfu/g respectively. Which is above the recommended limit of 10cfu/g. Is the a different method of doing dilution because it seems difficult getting the recommended limit after several analysis on different ice cream

Hello Terence,

I am hoping you can provide some additional details.  If the limit of your product is 10 cfu/g and you are failing that specification I think you need to ask "why?" you are failing that specification.  Making sure you are using aseptic technique, swapping out tips or pipettes between dilutions and plating, mixing well, are a couple of things you can look at in your process to ensure that you are not carrying over.  

A little clarity on the question would be helpful.

Best Regards,

Michael

Got an ice cream product which I made a 10x and 100x dilution and inoculated 1ml of each on coliform plate and after incubation I got 4 and 1 counts respectively. Meaning the cfu is 40cfu/g and 100cfu/g respectively. Which is above the recommended limit of 10cfu/g. Is the a different method of doing dilution because it seems difficult getting the recommended limit after several analysis on different ice cream

Hi Terence,

Note that counts of 1 and 4 are far too low for accurate evaluation and from a correlation POV simply don't make microbiological sense.

I also hope you are making duplicates at each "dilution".

Regarding the "dilution" you need to explain precisely/numerically what you mean by 10x and 100x since previous similar posts have run into confusion over this and subsequently miscalculated.

For example IMEX a typical procedure starts with 25g sample/225ml diluent giving mixture A. ie 1ml of A contains 0.1g of sample.

After that various actions are possible.

For example if you plated 1ml of A and got 20 colonies the meaning is that 0.1g sample generated 20 colonies which implies 200cfu/gram sample

this may give you some specific information to go on.......i daresay you've got a cleanliness issue somewhere

http://dairytechnolo...lity-assurance/

Hey Charles

Precisely this is what I did I measured 11g of ice cream and added to 99ml of dilution solution. This gave us a dilution factor of 1/10 (10x). From this mixed I transferred 1ml and inoculated it on 3m coliform plates and after incubation for 24hrs I counted 4 colonies. Calculating the cfu give us 40cfu/g I'm I doing the right procedure and calculation? The FDA recommeds 10cfu/g

Did you continue the dilution of the same solution to 1000?  Have you repeated the test again after putting all equipment that you'll be using through the autoclave?

http://dairytechnolo...lity-assurance/

Hey Scampi

No I didn't since dilution is purposely to reduce the number of bacteria and make bacteria count easier I didn't because the ice cream undergoes pasteurization during production process which kind of eliminate most coliform bacteria ....I was kind of right since I got only 4 counts from inoculating 1ml from dilution of 11g into 99ml

The purpose of additional dilutions is to add statistical balance to the testing. If you had 4 at 10x, you'd expect to see 40 at 100x......if you don't then you know you need to try again

Thanks Scampi for the link

Went through but the procedure wasn't using premade plates like 3m petrifilm plates and the recommended count was 15- 20 which I think may give us a cfu above 10cfu/g .... please clarify me more if I'm getting the whole dilution and calculation wrong

Basically this what I'm doing

I'm doing a coliform count on ice cream and the recommended level for coliform is 10cfu/g
So I weighed out 11g of ice cream and dissolved it in 99ml of solution giving us a 1/10 (10x) dilution. From this dilution I inoculated 1ml from the mix to 3m coliform plates and incubated for 24hrs and got 4 counts this gives us the cfu to be 40cfu/g . I did several analysis with different ice cream but couldn't get the recommended level of 10cfu/g my results was always above it .I also did a 100x dilution by dissolving 1g of ice cream to 99ml of dilution solution and carry out thesame inoculation procedure as 10x dilution and got 1 count giving us a cfu of 100cfu/ g Its seems im doing it the wrong way because I can't figure out what's wrong

Thanks Scampi

Hi Terence,

Sorry to be negative but IMO yr micro. interpetations are flawed. You are making a highly inaccurate calculation from an incomplete  procedure. 

In order to estimate such low cfu/gram values with more accuracy you need to either use a routine MPN procedure or adjust yr method of direct plating (involves using multiple [usually 2-3] plates).

With yr method, IMO only the 1st dilution has some meaning (eg a count of 0 or 1 is 'acceptable'). So, as you say, 4 colonies suggests out of spec.

But surely need more data to "confirm".

@ Scampi

Not sure if you meant (a) 4cfu/g or (b) total 4 colonies on the plate. For sufficient colonies on plate(s), (a) calculation should approx.stay the same, (b) total colonies on plate should decrease by approx 10x.

PS - I note the limit is slightly increased for non-plain ice cream.

PPS - No idea what method is officially prescribed, if any. i assume you are using VRB agar. Note that, IIRC, depending on the food matrix this can sometimes give false positive results for coliforms.

.

@ Charles

Let say I wish to continue using the direct plating method from your opinion

I need to use two or more plates

What about my dilution of samples do I need to adjust that .

If you have precise methodology of direct plating of ice cream will be happy to try it.

@ Charles and Scampi

Check this out
https://www.google.c...vf-gJk7idW7vf_Q

Also check out this video

Hi Terence,

There are a variety of methods to increase the sensitivity while attempting to maximise growth potential. .

For example, one official option I noticed (for S.aureus) is to use plating of 0.4ml from the first (1:10)_dilution on to each of 5 plates so total sample is 2ml = 0.2g. Now 1 observed colony = 5cfu/g. The arithmetic could equally apply to Coliform. Other permutations are possible. Another approach if practical is, as in attachment below, to modify the 1st dilution. This is usually limited by the necessity for clarity on the plates.

This Petrifilm method enables a sensitivity such that 1 observed colony = 1cfu/g

 Petrifilm - enumeration of Coliform.png   538.36KB   1 downloads

However I would have thought that traditional MPN methods still offer the most reliable, routine, option for this project albeit more work / a lot slower. IIRC a standard 3x3-tube system offers down to  < 3 cfu/g.

Both types of methodology seem to be in routine use for ice cream.

I love  love love the 3M system. Easy to use, and you can add a plate reader that will dump the data into excel for tracking purpose. It also eliminates the need for petri dishes et all.

I would highly recommend you call them. They also give out great support....specialists in their field with a wide spectrum of industry knowledge. You can easily teach a new person the system. One kind of plate has a built in well, the other comes with a tool to spread the sample easily. With the reader, it takes a photo of the plate as well, so you can recall the actual plate if someone questions your findings

I believe the countable range for the E.coli Petrifilm is above 15 cfu, so you will need to adjust your dilutions to put your counts into this range.  Range is 15-150 so I would maybe just do an undiluted sample if possible if you've already done 10^-1 and 10^-2. You'd have to call 3M to see if your ice cream formulation is compatible with the agar on the plate or if you would need to treat it in some way before inoculation. 

No way to confirm that your product is out of limits until you can get something within countable range, but sounds like you may have an issue with the counts that you've already gotten. 

I believe the countable range for the E.coli Petrifilm is above 15 cfu, so you will need to adjust your dilutions to put your counts into this range.  Range is 15-150 so I would maybe just do an undiluted sample if possible if you've already done 10^-1 and 10^-2. You'd have to call 3M to see if your ice cream formulation is compatible with the agar on the plate or if you would need to treat it in some way before inoculation. 

 

No way to confirm that your product is out of limits until you can get something within countable range, but sounds like you may have an issue with the counts that you've already gotten. 

Hi MsMars,

The problem in this case is at the low end due presumably to statistical error.

Direct sampling is OK if you can see the colonies and pipette/flow the liquid. Most ice-cream methods seem to start with 1/10. The above Petrifilm method uses  5:1 which I guess was as close as felt practical. Especially if not Vanilla.

These difficulties in sensitivity were the driving force for developing MPN methodology although even MPN estimates still have sizeable error bands. People love plating because it's quick (and dirty?).

Hi MsMars,

 

The problem in this case is at the low end due presumably to statistical error.

Direct sampling is OK if you can see the colonies and pipette/flow the liquid. Most ice-cream methods seem to start with 1/10. The above Petrifilm method uses  5:1 which I guess was as close as felt practical. Especially if not Vanilla.

 

These difficulties in sensitivity were the driving force for developing MPN methodology although even MPN estimates still have sizeable error bands. People love plating because it's quick (and dirty?).

Yes - perhaps also Petrifilm is not the best media for this type of testing if your limit is 10cfu/g and countable range is above that.  I definitely wouldn't base any type of COA on Petrifilm plating if your limit is below the countable range.  

I do agree - Petrifilm plating is quick (and definitely dirty!) with fairly easy methodology and interpretation. And I was thinking only of vanilla of course. How could I forget about chocolate ice cream? 

...am I the only one reading this and seeing someong trying to test their way into compliance?

Send it out to another lab to see what their results are compared to yours, but it sure as heck seems like you would be perfectly happy with the accuracy of your method had it provided results within specification. This is my main beef with "in house" micro, all negative results are perfect, all positive results must have been sampling contamination 

Sample additional ice cream and start looking for root cause of the contamination or failed pasteurization.

Hello Terence,

I am hoping you can provide some additional details. If the limit of your product is 10 cfu/g and you are failing that specification I think you need to ask "why?" you are failing that specification. Making sure you are using aseptic technique, swapping out tips or pipettes between dilutions and plating, mixing well, are a couple of things you can look at in your process to ensure that you are not carrying over.

A little clarity on the question would be helpful.

Best Regards,
Michael

Basically this what I'm doing

I'm doing a coliform count on ice cream and the recommended level for coliform is 10cfu/g
So I weighed out 11g of ice cream and dissolved it in 99ml of solution giving us a 1/10 (10x) dilution. From this dilution I inoculated 1ml from the mix to 3m coliform plates and incubated for 24hrs and got 4 counts this gives us the cfu to be 40cfu/g . I did several analysis with different ice cream but couldn't get the recommended level of 10cfu/g my results was always above it .I also did a 100x dilution by dissolving 1g of ice cream to 99ml of dilution solution and carry out thesame inoculation procedure as 10x dilution and got 1 count giving us a cfu of 100cfu/ g Its seems im doing it the wrong way because I can't figure out what's wrong

Hi Terence,

There are a variety of methods to increase the sensitivity while attempting to maximise growth potential. .

For example, one official option I noticed (for S.aureus) is to use plating of 0.4ml from the first (1:10)_dilution on to each of 5 plates so total sample is 2ml = 0.2g. Now 1 observed colony = 5cfu/g. The arithmetic could equally apply to Coliform. Other permutations are possible. Another approach if practical is, as in attachment below, to modify the 1st dilution. This is usually limited by the necessity for clarity on the plates.

This Petrifilm method enables a sensitivity such that 1 observed colony = 1cfu/g
Petrifilm - enumeration of Coliform.png

However I would have thought that traditional MPN methods still offer the most reliable, routine, option for this project albeit more work / a lot slower. IIRC a standard 3x3-tube system offers down to < 3 cfu/g.

Both types of methodology seem to be in routine use for ice cream.