20 replies to this topic

[Krparker]

Hi all, 

 

I am in the process of creating a product analysis schedule, but I cant seem to get through it and would like some advice from all you experts out there!!!   Background: We are a small, seasonal confectionery manufacturer (Most of our products are 90% flavored chocolate with an inclusion here or there). We are going for SQF L2 certification. Ingredients consist of chocolate, flavoring and inclusions (hard candy, cereal, salt etc)

 

Since we are a very small company, I am trying to keep the product sampling to a minimum as our budget is not huge, and the workforce for managing all these programs/ timelines is basically only me. Because of this I want to make a sampling schedule that takes into account the risk of the ingredient, the supplier status (certified, HACCP etc.), provision of COA's etc.. to minimize overall testing. 

 

My main question is do all samples and products have to be tested against the COA every year, or more than every year?... I would like to make sampling for ingredients that come from certified (GFSI) suppliers, with COAs for each lot tested at a frequency of 2 years (or more if acceptable). More of a validation that the COAs are representative. Tests without COAs would get tested upon receiving. Would this be an acceptable product analysis program to an auditor? 

 

Tests I would do for ingredients would be APC, Coliform, Salmonella & for finished product APC, coliform, yeast/mold.  We do regular quality sensory checks.  

 

Thoughts? Am I doing necessary testing by testing all ingredients? 

 

Thanks!

[Quality Ben]

Hi Krparker,

 

If you are trying to keep sampling / costs to a minimum then I would put the emphasis on the suppliers to give a COA with each batch etc. That is what we do with lots of products as it would be too costly and logistically ineffective to do it ourselves. If they are a good supplier and they wish to do business with you they will normally comply or at least come to a middle ground. The tests you are doing sound ok but have you thought about water activity? 

[Krparker]

Hm, yes water activity-- not sure how I didnt think of that one. 

 

Basically, i just want to make sure I am meeting SQF 2.5.6.1 Product sampling  "inspection and analyses are completed at regular intervals as required and to agreed specification".. If we are only testing our ingredients/ products once every 2 year, is this acceptable?

 

Also, a auditor posed to us how we know our COAs are actually correct/ true, and how often we validate against them. Once a year, twice, every five years? 

[Quality Ben]

Once every two years seems too lax to me but I do not work with confectionery. I think it is risk based...you may want to test every batch / product for a few months etc to build up data that you can then use to prove that a move to testing every year / 2 years is ok. As far as the auditor asking about COAs - you need to verify your lab. Are they NATA accredited etc? If they are a reputable lab and have accreditation then that is all you need. The auditor asking that seems fairly pedantic to me and I would not be too concerned...just do your research and have the data to prove reputability of the lab.

[fgjuadi]

Ah ha!  I am in a similar situation - We are a tiny manufacturer of chocolate with inclusions and I recently set up my raw material testing. 

 

I bought a few little instruments pH, Brix, moisture & fat analyzer, a small set of sieves with 10/20/30 mesh sizes, viscosity meter, had them calibrated,  created lot/tracking & lab results forms & SOPs for myself, and took proficiencies for them from an ISO accredited lab.  I test every ingredient coming in (we are tiny so this is easy, only 3-4 ingredient deliveries a week, testing takes minimal time) if these are part of the specs on the COA. 

 

I've found that 20% of the ingredients we were getting were out of spec!  AFTER my proficiencies! And calibration!  Mostly it was moisture content, and if I get something I go composite sample all the containers in the lot (this part is a pain in the patootie (ed note: I sanitized that for you cuz I keep it clean like the Foo Fighters), so make sure you have a plan for rejection, & records)  I haven't be testing that long and it may just be that we are getting dried fruit, which will vary, but..!  20%!  These instruments were inexpensive and have already saved us from overpaying for water more than they're worth, and more than it took to pay me for the time testing & documenting.

 

I started annual COA testing for pathogens on our high risk for salmonella ingredients - we use fresh mango, but mostly almonds and peanuts, and stuff from the field like macademia nuts and coconut, strawberries.  I'm considering going forward and testing COAs from suppliers that sent us materials that were out of spec, and testing things like our chili powder because there was just a recall for Salmonella tainted chili powder.  That step would have to be approved by the board of directors/senior management, and would mean we would be holding ingredients for longer before release, so it's a cost/risk to the business.  But the risk of not testing is too high - we don't have a kill step. 

[Charles.C]

Dear Krparker,

 

A few comments.

I deduce you are curently only referring to analyses of input ingredients, not finished products ?

 

Any intensive sampling scheme for the micro. items you mention is IMO going to rapidly strain an internal micro.lab unless you have unlimited funding, ie the frequency is likely limited by resources, not risk assessment / sampling theory. For many labs the number of samples per target item for routine lot evaluation is, rightly or wrongly, one. ,  particularly as the result is often only available after the ingredient has already been utilised. I think most auditors are well aware of such limitations. The relevance of traceability is clearly paramount.

 

A proposed input control scheme usually starts by developing yr own specifications. Evidently the safety ones would be prioritised if  we are talking about  a  FS system.

Any COA is logically validated as to (a) whether in compliance with yr own specifications  and (b) whether accurate or not. If the COA is provided by an appropriately certified lab, eg iso 17025, published statements exist to support your believing it, particularly if you lack the capability to self-analyse and/or complex entities are involved. Although it is presumably logical to do a "ring test" a few times per year for critical factors.

It is also crucial to have a follow up SOP in case that cross-checking does fail.

 

IMEX -

(1) many processors have sufficient experience/knowledge to approx. rank the risk of their inputs. If unknown initially use a moderate risk status / get data.

 

(2) many processors are forced to apply an ad hoc sampling/analysis scheme to inputs based on resources, eg lab can handle maximum x samples per day/week, this number is then apportioned over the incoming lots. This assumes a relatively stable incoming scenario. Exceptions require separate handling. And detected failures.

 

Rgds / Charles.C

[Snookie]

  For anyone making chocolate if you need organoleptic testing, I'm available.  Don't think I've met a chocolate I didn't like.        PICK ME PICK ME  !!

Edited by Snookie, 16 May 2014 - 09:06 PM.

[fgjuadi]

  For anyone making chocolate if you need organoleptic testing, I'm available.  Don't think I've met a chocolate I didn't like.        PICK ME PICK ME  !!

Where were you two weeks ago?  We just hired a consultant to redesign our sensory program & come in to train us :)

[Oraine]

similar situation here

[Snookie]

Where were you two weeks ago?  We just hired a consultant to redesign our sensory program & come in to train us :)

 

Ahhhhh.......so sorry I missed out.  Anytime you need a taste tester.....   After all chocolate is an essential nutrient and one of the four basic food groups   

[Krparker]

Thanks everyone for your replies!  I should also note that we only produce 5 months out of the year, so when I say "annually" it's really once a season, not for 12 months of product. 
 
Let me make sure I'm interpreting all the suggestions right.... I am mainly concerned with microbiological aspects.  
 Because the suppliers use accredited labs (I checked), the need to validate the COAs is unnecessary, except maybe as validation for high risk or sensitive items The remaining testing should allotted between ingredients that are high risk, and or/ not tested or, not from accredited labs, and regular ing testing if left in the budget.    
 
If that's the case then I wont be testing our flavorings, and most inclusions. The chocolate comes from  an accredited lab, but since its high volume and somewhat high risk, I will test against it at least annually including salmonella.. Not sure if that seems right to me. 

 

On the flip side, buying some quality- testing equipment is something  I will definitely be looking into!! 

[Charles.C]

Dear Krparker,

 

A (maximum) precise comment may inevitably depend on the precise text (or perhaps its auditorial interpretation) of the projected SQF certification standard.  ? (not a SQF user myself). L2 includes non-safety ? (I sincerely hope not ).

 

Rgds / Charles.C

[fgjuadi]

I am mainly concerned with microbiological aspects.  
 Because the suppliers use accredited labs (I checked), the need to validate the COAs is unnecessary, except maybe as validation for high risk or sensitive items The remaining testing should allotted between ingredients that are high risk, and or/ not tested or, not from accredited labs, and regular ing testing if left in the budget.    
 
If that's the case then I wont be testing our flavorings, and most inclusions. The chocolate comes from  an accredited lab, but since its high volume and somewhat high risk, I will test against it at least annually including salmonella.. Not sure if that seems right to me. 

 

On the flip side, buying some quality- testing equipment is something  I will definitely be looking into!! 

 

 

If you're only concerned with micro, the chocolate isn't high risk (usually water activity is very low), flavorings aren't high risk - Things that are high risk for you are probably nuts or stuff that comes from a field if it isn't processed like fruit. 

[Charles.C]

Dear Magenta_m,

 

I

f you're only concerned with micro, the chocolate isn't high risk (usually water activity is very low), flavorings aren't high risk

 

 

Yes and No.

 

Relatively old link but AFAIK, still relevant -

 

http://www.gewponoi....N_5,171-175.pdf

 

Rgds / Charles.C

 

PS - and one from bitter experience -

 

http://www.foodprote...erlin/Banks.pdf

[fgjuadi]

Dear Magenta_m,

 

I

 

Yes and No.

 

Relatively old link but AFAIK, still relevant -

 

http://www.gewponoi....N_5,171-175.pdf

 

Rgds / Charles.C

 

PS - and one from bitter experience -

 

http://www.foodprote...erlin/Banks.pdf

Touche.  That presentation is pretty nice for basic overview of how to make chocolate.

I was assuming the OP was working from couverture, not cocoa beans,  & had standard GMPs in place.

The FDA does list chocolate as high risk for Salmonella as well -

 

http://www.fda.gov/f...k/ucm069966.htm

 

But Food Safety.gov and the CDC do not -

http://www.foodsafet...ella/index.html

Edited by magenta_majors, 21 May 2014 - 10:54 AM.

[mruth84]

We make sure our vendors are audited under a GFSI scheme, and accept each lot based upon COA.

[Charles.C]

Dear magenta_m,

 

Rather

 

Interesting links, thks.

 

One fundamental difficulty, not for the first time, is due to terminology. There is no uniform definition of a "high risk product". For example, here are “definitions” from the glossaries of 2 well-known food standards.

 

SQF 7.1

A  food  that  may  contain  pathogenic  microorganisms  and  will  support formation  of  toxins  or  

growth  of  pathogenic  microorganisms,  and   has  a  significant  likelihood  of  growth  causing  

illness  or  injury  to  a  consumer  if  not  properly  produced,  processed,  distributed  and/or  

prepared  for  consumption.    It  may  also  apply  to  a  food  that  is  deemed  high  risk  by  a  

customer,  declared  high  risk  by  the  relevant  food  regulation  or  has  caused  a  major  

foodborne  illness  outbreak

 

 

BRC6

A chilled ready-to-eat/heat product or food where there is a high risk of growth of pathogenic micro-organisms.

 

 

 

The US FDA also seem to be in somewhat of a state of flux on this topic -

 

http://www.fdalawblo...risk-foods.html

 

Rgds / Charles.C

[fgjuadi]

Dear magenta_m,

 

Rather

 

Interesting links, thks.

 

One fundamental difficulty, not for the first time, is due to terminology. There is no uniform definition of a "high risk product". For example, here are “definitions” from the glossaries of 2 well-known food standards.

 

SQF 7.1

 

BRC6

 

The US FDA also seem to be in somewhat of a state of flux on this topic -

 

http://www.fdalawblo...risk-foods.html

 

Rgds / Charles.C

Perhaps off topic

 

Though questioning how often to validate a COA based on micro risk would mean discussing if Chocolate is high or low risk for Salmonella contamination seems pretty in line with the OP.

 

And how do we determine high risk?  So, this topic flows

 

Ugh, that first SQF sentence sounds like every food to me.  Anything you will eat will support growth of life. 

BRC makes more sense - if you must refrigerate and there's no kill step, its risky, cuz you might leave it out in the sun too long.

 

The FDA are real optimists - the include Economic impact.  I guess if your food only kills the poor it's good to go, so long as the ones that shop survive & are rendered so dull they have brand loyalty.  Though I like the frequency of illness, death, and growth potential being taken into account.

 

What I really want is a list by pathogen of what is considered high risk, which you would think every standard would have as an appendix in the back thats regularly updated.  One thing about the non-GMO standard I really like is that their high risk category is pretty clear and a list of ingredients which are high risk.  Of course, this is probably much easier because GMO strains are heavily cataloged and finite, unlike the endless possibilities present in food handling.

[Charles.C]

a list by pathogen of what is considered high risk

 

 

Actually a lengthy  3-tier risk list was issued around 2 decades ago (USA, seafood haccp origin from memory). Has been posted here 2-3 times. I think a similar concept is also built into the well-known ICMSF micro. sampling tables for lot accceptance.

 

But, although scientifically based, such lists have never really caught on within the risk assessment genre, probably because most users prefer a simple high/low approach for material input evaluations,eg the published haccp input trees.  Similarly, in risk probability calculations in hazard analyses,  "pathogens" tend to be auto-labelled as high severity, regardles of relative potency. i guess it matches the 3x3 matrix.  

 

Rgds / Charles.C

[Snookie]

The FDA are real optimists - the include Economic impact.  I guess if your food only kills the poor it's good to go, so long as the ones that shop survive & are rendered so dull they have brand loyalty.  Though I like the frequency of illness, death, and growth potential being taken into account.

 

Thank you!

Edited by Snookie, 22 May 2014 - 07:17 PM.

[Charles.C]

Dear magenta_m,

 

One thing about the non-GMO standard I really like is that their high risk category is pretty clear and a list of ingredients which are high risk.  Of course, this is probably much easier because GMO strains are heavily cataloged and finite, unlike the endless possibilities present in food handling.

 

Oh Dear!

 

If you apply zero tolerance, yes the category is fairly well defined. If otherwise, it's back to quantitative risk assessment.

 

Rgds / Charles.C