9 replies to this topic

[GBCones01]

We have a baked product whose process has a validated kill step achieving over a 5-log reduction. One of our products is shipped to Canada so we send it for APC, Coliform and Y&M testing before release and shipment. Recently, we have been struggling with out of spec APC results (spec limit is <1,000 which I believe is too low anyway). The main issue is the inconsistency with which these out of tolerance results occur. We have swabbed all our equipment used to make this product and it was all completely fine.  I thought maybe it was the flour, despite the validated kill step, as I had been involved in APC discussions at another company where we were advised by suppliers that crop year and other crop variances can and will affect APC levels. I thought I confirmed this by tracking our flour deliveries against the lots of product being tested. It looked like once we started on a new load of flour the results evened back out to <10 consistently. Suddenly we have extremely high and inconsistent APC results again. Monday was 200,000, Tuesday was <10 Wednesday was 250,000.  I can't seem to make sense of it and I have run out of ideas as to the root cause and how to stop this. I'm hoping others may have some insight. Until these past 3 results the out of spec occurrences were nowhere near this high ranging from 86,000 to 1,900. Any ideas, help or guidance would be appreciated - I'm not a microbiologist so I've reached the capacity of my knowledge here. 

 

 

 

 

[TimG]

From one non-micro guy to another, are you confident there is no contamination during testing? 

[Scampi]

It actually sounds like the contamination is "happening" in the flour bin

 

e.g. it may not be stored in 100% the right conditions, allowing the APC to change over the storage time of a lot of flour

 

You are sending finished goods IN their final packaging for testing?????

[GBCones01]

No - I wouldn't be confident enough to say the lab isn't causing some sort of contamination. They are accredited, but there have been separate issues where they later stated that some presumptive positives we were getting were their fault. Unfortunately. 

[GBCones01]

It actually sounds like the contamination is "happening" in the flour bin

 

e.g. it may not be stored in 100% the right conditions, allowing the APC to change over the storage time of a lot of flour

 

You are sending finished goods IN their final packaging for testing?????

 

I would possibly agree with you on the APC changing over the storage time, and maybe it is, but we are not going through a super sack per day/lot on this product, so when we see wild fluctuations over just 3 days it doesn't make any sense - all of those lots would have been on the same super sack of flour. 

Does a validated kill step not have an effect on APC counts in flour? 

 

The product that we are sending for testing is a composite sample across all three shifts put into one smaller sample bag. The sampling method at the plant would have everyone gathering samples to wear gloves. 

[jpranav23]

GBCones01 

If the raw material (flour is really bad then you can have variation ) however to achieve 200000 at a 5 log reduction the raw materials would be worst . 

 

I will suggest following approach 

1. Simultaneous Sampling from one run - flour , dough , end of oven , end of cooling , packing material and post packing . Draw a trend 

2. Check if there are any uncleared deposits in flour , mixer which is partly getting released incresing sudden load > In liquids biofilms impacting results is a known fact. 

3. Look at the cooling , dehumidifier storage - is there condensation observed leading to growth and cross contamination 

4. 

We have a baked product whose process has a validated kill step achieving over a 5-log reduction. One of our products is shipped to Canada so we send it for APC, Coliform and Y&M testing before release and shipment. Recently, we have been struggling with out of spec APC results (spec limit is <1,000 which I believe is too low anyway). The main issue is the inconsistency with which these out of tolerance results occur. We have swabbed all our equipment used to make this product and it was all completely fine.  I thought maybe it was the flour, despite the validated kill step, as I had been involved in APC discussions at another company where we were advised by suppliers that crop year and other crop variances can and will affect APC levels. I thought I confirmed this by tracking our flour deliveries against the lots of product being tested. It looked like once we started on a new load of flour the results evened back out to <10 consistently. Suddenly we have extremely high and inconsistent APC results again. Monday was 200,000, Tuesday was <10 Wednesday was 250,000.  I can't seem to make sense of it and I have run out of ideas as to the root cause and how to stop this. I'm hoping others may have some insight. Until these past 3 results the out of spec occurrences were nowhere near this high ranging from 86,000 to 1,900. Any ideas, help or guidance would be appreciated - I'm not a microbiologist so I've reached the capacity of my knowledge here. 

[Scampi]

to be clear, you're NOT sending in actual finished goods in the final packaging?????  Post kill step???

 

 

The lab can make a composite of whatever you send them

 

 

Have you done ACC settling plates to see if it is your air contaminating your products?

[GBCones01]

GBCones01 

If the raw material (flour is really bad then you can have variation ) however to achieve 200000 at a 5 log reduction the raw materials would be worst . 

 

I will suggest following approach 

1. Simultaneous Sampling from one run - flour , dough , end of oven , end of cooling , packing material and post packing . Draw a trend 

2. Check if there are any uncleared deposits in flour , mixer which is partly getting released incresing sudden load > In liquids biofilms impacting results is a known fact. 

3. Look at the cooling , dehumidifier storage - is there condensation observed leading to growth and cross contamination 

4. 

Thank You! This gives me some more areas to consider and look at. I wasn't necessarily looking at the mixer because the batter is cooked/goes through that kill step. 

[GBCones01]

to be clear, you're NOT sending in actual finished goods in the final packaging?????  Post kill step???

 

 

The lab can make a composite of whatever you send them

 

 

Have you done ACC settling plates to see if it is your air contaminating your products?

We are sending in our "final product" to the lab for testing - the material that will eventually go to the customer. So yes, it is post kill step (oven/baking). 

 

We are compositing our shifts into one sample for the lab and I am not having them composite lots from there so that we can have results on each individual lot - otherwise it's impossible to know which lot(s) are causing the out of spec result. 

 

I have not done air plates as the problem is not with Y&M and the completely sporadic results didn't necessarily point me to think air quality - but perhaps it's something to reconsider. 

[GMO]

From having worked in bakery, I once found a production supervisor using raw flour to "lubricate" the line to ensure the bread flowed.  

 

So I'd follow through like any investigation in your plant for micro.  Look and probably test at different stages in the process.  Sample yourself if you need to.  I agree that anything post bake is going to be something you need to investigate thoroughly.  Everything from any condensation to the cleanliness of fans used to circulate air.  Even evaporator units can be part of the problem here if you don't regularly clean and disinfect them.  

I'd get a factory plan and be systematic about it.  Swabs, samples, hands, air, packaging etc. Also don't rule out your sampling process.  If your compositing your sample, you're opening the bags.  

 

But one thing I'd also recommend is looking at different times.  Sadly you're doing a composite so the data won't point you in the direction of a certain shift but much like my "flour lubrication" example, you can find odd practices which impact bacterial loading that only happen at certain times.