9 replies to this topic

[jkarau]

Hello!  I am looking to start a testing program where I look for allergens at incoming receipt of spices and herbs (i.e. cinnamon, garlic, onion, paprika, cumin, etc.).  I've determined that laminar flow test strips are the best option for this type of testing.  I have an idea of which allergens I want to test for as well as which incoming materials I want to test.  The plan will include testing from every supplier who provides the selected incoming materials.

 

Has anyone implemented this kind of allergen testing protocol?  If so, what did you do and has it been accepted by external auditors?

 

Any and all help/suggestions are welcome!

 

Jacqui

[GMO]

I actually think this is a great idea as contamination with allergens in spices has happened and more than once.

 

First question you might want to answer is risk of cross reactivity on those strips.  I'd get some advice on that.  Also do you then follow up any positives with ELISA to confirm?

 

But my second question may make you rethink (or may make any of your customers furious with you).  What will you do with the results?

 

Now I actually hate myself for asking this question because it's all about business protection and NOT about food safety.  But in the UK we have a retailer called M&S and they do not let you test ANYTHING for allergens (ingredient or product) unless it's in your control AND you have their approval.  Why?  Because it might not impact only your company.

 

Think about it.  You have a delivery of ground cumin for example.  That same batch is also delivered to a meals manufacturer, a spice paste company, a fast food restaurant, the sandwich supplier etc etc.  You find peanut protein in it.  Imagine the size and impact of recall...  Ooh I feel dirty talking about commercial considerations.  

 

That's why I hate myself because if it's there I don't actually care how much money is wasted.  BUT if you trust your supplier, would it be better if they did the testing before any of that batch left their control?  

 

Or if we flip this on its head, TELL your supplier you're going to start doing this testing at intake and I guarantee they will start testing every batch before it leaves their control because of the implications. 

 

Ok I've talked myself back into it.  Crack on but make sure everyone knows then on their head be it!  

[Tony-C]

Hi Jacqui,

 

You might find the following interesting reading, both are from the Spice and Seasoning Association:

 

Allergen Risk Assessment Model for Dried Herbs and Spices

 

 allergen-risk-assessment-model-for-dried-herbs-and-spices.pdf   365.58KB   8 downloads

 

Guidance on Authenticity of Herbs and Spices - Industry Best Practice on Assessing and Protecting Culinary Dried Herbs and Spices

 

 guidance-on-authenticity-of-herbs-and-spices.pdf   6.19MB   7 downloads

 

Kind regards,

 

Tony

[GMO]

Thanks for linking these Tony.  Have you read through them? 

 

Interesting they're advocating using a reference dose in the UK and they even are suggesting reference doses in recipes which are (as a home cook) woefully short.  For example their "standard" recipe for use of smoked paprika is hummus (really?!) and 1/2 tsp in 4 portions.  Anyone who has made pulled pork, barbecue sauces, patatas bravas or even my bloody awesome chilli, will use FAR more than that.  And 1 tbsp of garam masala across 4 portions?  I think I'm learning I'm heavy on the spices...  

 

It would be interesting for a company to do that without contacting the FSA, if there was still an allergic reaction (as susceptibility varies) I'd worry they would be duly diligent if they'd not discussed this before leaving in market.  ESPECIALLY considering the risk to other companies and the fact they could be doing their own, say, cleaning validation testing (including next product) oblivious to known contamination in an ingredient.  

[jfrey123]

Testing the incoming spices is going to be tricky to have confidence in the efficacy.  It's one thing if you're concerned about them having undeclared add-ins from a fraud perspective, but catching the incidental cross contamination from a poor cleanout is akin to finding a needle in a haystack.  How much spice does your company order in an average load?  Couple boxes here and there could be manageable, but if you're receiving truck loads with 20-25 cases per pallet, you're not going to be able to effectively test that load.  You'll be pulling samples at random, and anyone auditing this process for you is going to want to poke holes at your determined frequencies/decision trees on how you quantified the number of samples:  One case from each pallet?  One case from every other pallet?  Is 5 samples total sufficient?  Why not 20 samples then batched?

 

As for which allergens, I suggest milk, wheat, and soy.  Those three are common in the pre-made spice rubs you can purchase for BBQ or whatnot, and if your spice supplier blends and packages such pre-made rubs there is a risk of poor sanitation leading to cross-contamination.

[jkarau]

Testing the incoming spices is going to be tricky to have confidence in the efficacy.  It's one thing if you're concerned about them having undeclared add-ins from a fraud perspective, but catching the incidental cross contamination from a poor cleanout is akin to finding a needle in a haystack.  How much spice does your company order in an average load?  Couple boxes here and there could be manageable, but if you're receiving truck loads with 20-25 cases per pallet, you're not going to be able to effectively test that load.  You'll be pulling samples at random, and anyone auditing this process for you is going to want to poke holes at your determined frequencies/decision trees on how you quantified the number of samples:  One case from each pallet?  One case from every other pallet?  Is 5 samples total sufficient?  Why not 20 samples then batched?

 

As for which allergens, I suggest milk, wheat, and soy.  Those three are common in the pre-made spice rubs you can purchase for BBQ or whatnot, and if your spice supplier blends and packages such pre-made rubs there is a risk of poor sanitation leading to cross-contamination.

Yes, this is the proverbial "sticky wicket"- how to justify the sampling/testing protocol we use.  Hypothetical example:  we receive 11 pallets of spice (20 containers per pallet, 220 total containers)- do I sample one container per pallet?  Sample and test the square root of n containers?  Composite those samples or run each individually?  Test just one container? 

 

I also have to keep my costs down, so if each test for each allergen costs $20 and I test for 5 different allergens per receipt, based on the hypothetical example I could spend anywhere from $100 up to $1100 for just one receipt of spice.  

 

I am looking at milk, soy, wheat/gluten, sesame, and mustard.  I'm open to including peanut in this program, based on the cumin issue about 10 years ago.

 

We plan to do verification/proof of concept testing when we get our kits in: we will take a product we know has the allergen of concern in it and a product we know does NOT have the allergen of concern as a control, prep both per kit instructions, and document the results.  We are planning to use white sugar as the control: sugar companies only handle sugar, so the risk of incidental allergen cross contact is minimal.

[jkarau]

I actually think this is a great idea as contamination with allergens in spices has happened and more than once.

 

First question you might want to answer is risk of cross reactivity on those strips.  I'd get some advice on that.  Also do you then follow up any positives with ELISA to confirm?

 

But my second question may make you rethink (or may make any of your customers furious with you).  What will you do with the results?

 

Now I actually hate myself for asking this question because it's all about business protection and NOT about food safety.  But in the UK we have a retailer called M&S and they do not let you test ANYTHING for allergens (ingredient or product) unless it's in your control AND you have their approval.  Why?  Because it might not impact only your company.

 

Think about it.  You have a delivery of ground cumin for example.  That same batch is also delivered to a meals manufacturer, a spice paste company, a fast food restaurant, the sandwich supplier etc etc.  You find peanut protein in it.  Imagine the size and impact of recall...  Ooh I feel dirty talking about commercial considerations.  

 

That's why I hate myself because if it's there I don't actually care how much money is wasted.  BUT if you trust your supplier, would it be better if they did the testing before any of that batch left their control?  

 

Or if we flip this on its head, TELL your supplier you're going to start doing this testing at intake and I guarantee they will start testing every batch before it leaves their control because of the implications. 

 

Ok I've talked myself back into it.  Crack on but make sure everyone knows then on their head be it!  

 

What would we do with the results?   Fail/reject the material, which will cause lots and lots of issues up and down stream- but it will protect us. 

 

Would I like my suppliers to do this type of testing and provide results?  Absolutely, and I'm going to let my supply chain team handle that piece.  

 

Would that eliminate the need for testing on my end at incoming/intake- no.  Since the FDA considers allergen a SAHCODHA hazard (serious adverse health consequence or death to humans or animals) I cannot rely solely on supplier provided test results, I have to routinely verify the data my suppliers provide.  

 

So. Much. Fun!

[G M]

Beyond the sampling problem you could have a lot of cross-reactivity giving you false positives. 

 

We've seen it happen with spice blends where we know the supplier is grinding whole spices.

[GMO]

Beyond the sampling problem you could have a lot of cross-reactivity giving you false positives. 

 

We've seen it happen with spice blends where we know the supplier is grinding whole spices.

 

To be fair that was mentioned in the references Tony shared.  Not spices but there was a famous case with wheat from Italy where it was strongly suspected of cross reactivity between mustard and oil seed rape.  It caused a lot of bother I can tell you.

[emportllc]

We plan to do verification/proof of concept testing when we get our kits in: we will take a product we know has the allergen of concern in it and a product we know does NOT have the allergen of concern as a control, prep both per kit instructions, and document the results.  We are planning to use white sugar as the control: sugar companies only handle sugar, so the risk of incidental allergen cross contact is minimal.

Lots of other good points and questions already raised. A few more just in case helpful— 

 

• For your positive control — are you planning to test the product with the known allergen as-is? You might want to use it as a spike instead, mixed into another matrix. This should do two things:
  • lessen the likelihood of overloading your test kit, if it's a kit that's subject to hook effects
  • give you some added confidence regarding possible matrix effects / false negatives
• When you come across a possible cross-reactivity, consider having a backup kit brand or two in mind. Sometimes switching antibodies solves the issue.   
• "Fun" sugar anecdote, we did once have a bakery client with significant gluten come in via sugar. The vendor told them it was impossible, because the sugar came from a country that did not export wheat.