11 replies to this topic

[Marshenko]

Looking for tips on writing a policy and doing risk-assessment for not using metal detection on most of our packaging lines. This is a final step in completing all necessary programs to gain SQF Level 2 certification.

Background
We are a very small meat processing company producing portion control whole-muscle steaks and roasts. These steaks are cut from whole sub-primals which are purchased from approved suppliers.
Our three primary packaging lines then cryovac these steaks etc. Currently we are not using a metal detector.
We do not use needling as part of our process. All items are hand-trimmed, and whole sub-primals may be cut down using a vertical band-saw.
Bench trim generated as part of this process is used for ground beef. A "flip-flop" vacuum machine is used for packaging these items.
We currently own one 10 year old metal detector, which is a pain-in-the-ass to set up and change for different items.

The Proposed Plan
The whole muscle items (again, which are not needled) have a very-low risk of metal contamination that would not be visible. There have been zero instances of metal contamination complaints in the 20 year history of the facility.

My plan consists of:
1. The use of a metal detector on the two high-risk items that we do produce (ground beef, shaved steak). All ground/shaved high-risk product would go through the metal detector.
2. Hourly blade (both knives and saw) checks to ensure that no blades have been broken, or saw teeth missing. In the event that everything is good, low-risk product would not go through the metal detector. In the event that a non-conformance is found during one of these hourly blade check procedures, all product cut and packaged from the time of the previous good check until current would be placed on QC Hold and not released until it was sent through the metal detector. In addition, visual inspection of all benches/product contact surfaces currently in use would take place to ensure no metal pieces were detected.

Given the risk assessment of the product and the policy outline set forth .... what do you think?

[Charles.C]

Dear Marshenko,

I hv little experience of the US (beef/poultry) meat industry but I always get the impression that every safety aspect has been put under so many microscopes that every haccp-relatable facet has been multiply categorised for posterity.

If so, from my experience of auditors, such existing info. will be their first source of representative guidelines. If you match such expectations, usually plain sailing, if not be prepared for lengthy arguments. And I suspect MDs are nowadays a “cause celebre” particularly as the availability/cost is favourable compared to a decade ago (especially 2nd hand market).

I worked under similar constraints as yourself in a non-“meat” industry with a daily unpredictable MD for several years. The auditor’s patience with a process line that was out of step with every similar competitor’s finally ran out when the fda published a brief listing of certain types of equipment for which they would “automatically” assume that metal contamination was/was not an unavoidable possibility.

Hopefully no similar listing in yr business arena. Only plus was a new (good 2nd hand ) MD in 2-3 months .

Rgds / Charles.C

[George @ Safefood 360°]

The fact remains that even with the most robust control measures in production metal contamination can and does occur. Processes such as cutting, mincing and mixing will introduce the potential risk of contamination.

Retailers such as Tesco will almost certainly require systems that prevent metal contamination backed up by detection methods such as in-line metal detectors or x-ray. There is no doubt in my mind that certain retailers would not accept your proposed action plan arising from your risk assessment even though it appears you have put a lot of thought in to it and it sounds reasonable. The next question you need to ask is whether a SQF auditor will accept it? The quick answer is I don’t know. But it will almost certainly depending on what validating documentation you can offer and even then I am not convinced it will fly.

The difficultly is this. The basic philosophy of food safety management is the reduction of the risk associated with identified hazards (not necessarily their elimination). Metal, as a hazard, is one where the risks are clear. Mechanical processes which bring metal in contact with food with the intent of changing its shape, size etc. will introduce metal contamination. Good practices will definitively reduce the risk. However good practices coupled with detection methods will reduce it even more! If you can see my point? This is how the retailers will view it and I suspect many auditors. The only exception is where your risk assessment is water tight and the argument I present above may provide slight leakage. The greatest weakness I can see in your proposal is that it depends heavily on visual inspection of blades, teeth etc. to trigger a metal detection step. If this is your control then it might suggest you need in line metal detection to begin with – I’m just following the logical that some auditors may take.

A few things to consider:

Put together a detailed report on your complaint history for metal for those low risk products.

Conduct and record metal detection on those products that you intend NOT metal detecting. Cover a number of batches under typical production conditions. This will ‘prove’ that your assessment is correct.

Have in place a good training program for those doing the visual inspections.

Have a detailed procedure to back up damage to blades andthe follow up action. Consider how all this impacts on your RA in your HACCP plan. Are damaged blades handled as a CCP?

Consider doing a reconciliation of band saw blades. Weigh the blade before and after a number of production shifts and see if metal is being lost and if this is consistent with the visual inspection program. Again this more validation for your risk assessment.

These are just a few things that spring to mind but there may be more.

George

[MD & X-ray man]

There are a lot of good comments above, but there's one key area that seems to have been missed...

Before December 1941, the USA had enjoyed about 150 years without being attacked by a foreign power. So did that mean that the risk of attack was zero? Perhaps you'd like to ask the people of Pearl Harbour?

Just because you have not had a customer complaint for a given period, does that mean that there has not been a contamination, or that there will not be a contamination?

It may surprise you but when you sell / supply products to someone they (your customer) does not want your assurance that you / your process have not added metal to the product.

Rather, they want your assurance that you have taken steps to locate and remove ANY and ALL (possible) metal from the product, whether it was introduced by you or not. In other words, you have been looking internally, but what about metal that was already in the meat when it arrived at your premises?

I can tell you from first hand experience that there are lots of pieces of metal in meat when it arrives at the abattoir. In particular lead (bullets), sections of veterinary needles, wire etc. It seems from your comments that you are thorough and logical. Now apply that thought to external factors; how will you detect and remove contaminants that were in the meat before it arrived at your premises? How confident are you that your suppliers have removed these contaminants before they sent the product to you? How confident are you that your suppliers have not added contaminants to the products through their processes?

I hope this helps.

[Charles.C]

Dear MDX man / Marshenko,

Thks for the interesting post. I have no idea if (statistically) Australia is hazard comparable to USA.? Do you also have a comparable Pathogen scenario (my guess is that nobody does, bar Canada perhaps)?

If taking a more overall view, one might inject at least 3 more factors, ie prerequisites, economics, farm-to-fork safety philosophy.

This 2003 article quite rationally describes some claimed top physical hazards in USA, rating ( scientifically and emotively) bullets / needles as one of the 12 Dirty Dozen for physical nasties.
Its final paragraphs, (although measured) do seem somewhat conciliatory (pragmatic?) in tone.
 Foreign materials, the Dirty Dozen.pdf   89.55KB   367 downloads
FSIS ( and similarly other countries) seem to have focused on the third concluding factor listed in above ref. , ie as part of their farm-to-fork strategy (1996).
(http://www.fsis.usda...ubs/93-016F.htm )(see pg 38810 et seq.)

More specifically to the present thread, this (2011) FSIS directive seemed to me to offer some optimism for Marshenko's approach (perhaps depending on the precise process details) –

C. Establishments are required to consider food safety hazards in three broad groups
(biological, chemical, and physical) that may occur before, during, and after entry into
the establishment (9 CFR 417.2(a)(1)).

1. Examples of biological hazards include bacteria, viruses, and parasites;

2. Examples of chemical hazards include chemicals prohibited in foods or beyond
limits allowable in foods, toxins of microbial origin, allergens, specified risk
materials (SRMs), pesticides, antibiotics, and hormones; and

3. Examples of physical hazards include glass, wood, metal fragments from
equipment, needles, bullets, shot, wires, clips, and twist ties.

D. Establishments need to support the decisions they make in their hazard analysis.
An establishment may use various forms of scientific information to support that a
hazard is not reasonably likely to occur, and that a prerequisite program will prevent a
hazard from being likely to occur.

 fsis usda 5100.1Rev3.pdf   360.49KB   202 downloads
And similarly this one (2007, the Ohio State website is directed for current info.)
 supporting documentation for haccp decisions, Folk 2007.doc   1.46MB   277 downloads
I also noticed this (get-out) text in an earlier haccp plan for receiving/ processing steps respectively–

Physical: We acknowledge that some carcasses and meat products may occasionally have
buckshot, bullets, needles, etc. However, these cannot be identified at the time of
receiving and may not be found as the product moves throughout the process. There have
been no reported incidences of these physical hazards from Jan. 1, 1998 to Dec. 20, 2002.
We also considered that size and shape of metal will impact whether or not an object is a
food safety hazard according to Olsen (1998), “… classifies hard or sharp objects over
7mm in length as potentially hazardous while objects that measure between 2 and 7 mm
are normally considered a non-hazardous defect.” [See supporting documentation.]
Therefore, physical hazards are not identified as a reasonably likely to occur food safety
hazards for this process step.
We are addressing lead shot as a potential chemical hazard.
AND
Physical: We considered the potential for metal from the equipment as a physical food safety
hazard. There have been no reported incidences of these physical hazards from Jan. 1,
1998 to December 20, 2002. We also considered that size and shape of metal that may be
generated from the equipment to determine whether or not an object is a food safety
hazard according to Olsen (1998), “… classifies hard or sharp objects over 7 mm in
length as potentially hazardous while objects that measure between 2 and 7 mm are
normally considered a non-hazardous defect.” [See supporting documentation.]
Therefore, physical hazards are not identified as a reasonably likely to occur food safety
hazards for this process step.

 haccp decision-making documentation 2003.pdf   120.96KB   247 downloads
(eloquent but somewhat “gilding the lily” perhaps?)

HACCP can sometimes appear a hard-nosed (risk) methodology since from a legislatory POV, health statistics (aka “Incidents”) tend to dictate what are mandated immediate necessities (eg the Pathogen Rule). IMO economic issues are always involved somewhere, even if not explicitly. I recall the outcry from SME seafood processors in USA when HACCP was first imposed as a legislatory demand. Particularly disliked were the recommendations to install metal detectors ( “un-necessary and economically ruinous” ). The “necessity” was quickly pointed out as including possible alternatives, similar to the OP of this thread and the protests diminished. This situation is probably still relatively unchanged.

No doubt Marshenko is well aware of the above factors if they are still relevant .

Rgds / Charles.C

PS it's OT but just in case you are also handling deer meat, this may be of interest -
http://www.nrahunter...LeadIssues.aspx

as against -
http://www.scientifi...-ammunition-ban

[Barrie@RJT]

For me, a very interesting question and resulting thread - as I am the Technical Manager in a relatively small abattoir/cutting plant, similar to Marshenko's situation, battling with the same issue: a 15-year-old MD (metal detector, not managing director.....!) that is not the easiest to set up and operate!

Up until now, we had BRC Issue 5 for just 'BQAP' pigs, and as these went out as whole or part carcases, it was accepted that MD was not appropriate. However, we are now working towards Issue 6 and so cannot exclude all the other products we produce - and it's a hell of a lot, from steaks, joints, diced meat, to mince, sausages, burgers and sliced bacon. So from previous threads on this forum, I have resigned myself to the practical difficulties of operating the ancient MD for all the packaged/boxed products before they are dispatched.

One point, though - we have similar (but not as rigorous!) additional controls in place to Marshenko and I am treating the whole of our 'metal contamination control' as an OPRP rather than defining the MD as a CCP, which I feel much more comfortable with, given the age of our MD!

[Marshenko]

Thank you all for the informative responses.

I am going to have to move forward with this really ... due to time constraints and the impossibility of running all of our product through our existing metal detector.
We have a pre-assessment next week, and our site audit is scheduled for the first week of February. I can certainly update this thread once the auditor reviews our products and the program.

In addition to the verification program, the metal detection of all high-risk items ... would it be appropriate to also implement a random sampling of low-risk product at designated intervals as an additional validation step?

Also, FWIW ... the USDA has never said a thing about metal detection on our products.

Edited by Marshenko, 13 December 2012 - 10:49 AM.

[Charles.C]

Dear Marshenko,

I have attached the official USFDA viewpoint on (haccp) handling of metal contamination from process equipment within the seafood industry which IMO is a very nice compendium detailing the haccp handling of most of the usual options. It seems to me that, equipment-wise, the seafood/meat industries should not be that much different. The content seems to offer some support for the proposed logic in yr OP.
 metal inclusion, haccp FDA seafood Ch20.pdf   162.1KB   254 downloads

After a quick look, I did notice that it seems not difficult to find current USA meat haccp plans which seem to have no interest in metal detectors, eg this one from meathaccp.wisc ... -
 rawGround_flow_diag 05-03-12.doc   33KB   160 downloads

Good luck!

Rgds / Charles.C

[Marshenko]

Dear Marshenko,

I have attached the official USFDA viewpoint on (haccp) handling of metal contamination from process equipment within the seafood industry which IMO is a very nice compendium detailing the haccp handling of most of the usual options. It seems to me that, equipment-wise, the seafood/meat industries should not be that much different. The content seems to offer some support for the proposed logic in yr OP.
 metal inclusion, haccp FDA seafood Ch20.pdf   162.1KB   254 downloads

After a quick look, I did notice that it seems not difficult to find current USA meat haccp plans which seem to have no interest in metal detectors, eg this one from meathaccp.wisc ... -
 rawGround_flow_diag 05-03-12.doc   33KB   160 downloads

Good luck!

Rgds / Charles.C

Without seeing the rest of the Hazard Analysis, I have no doubt that that University of Wisconsin HACCP plan has grouped metal detection into packaging, which is why they have it listed as a CCP...

[Charles.C]

Without seeing the rest of the Hazard Analysis, I have no doubt that that University of Wisconsin HACCP plan has grouped metal detection into packaging, which is why they have it listed as a CCP...

Dear Marshenko,

Hazard Analysis attached. No MDs. B= Biological.
No suggestion of physical hazards at receiving. Maybe they are relying on history / quality supplier etc / eyesight.

For grinding, the text implies that the equipment is visually monitored "in-process" and also via the "cleaning" prerequisite. Result > hazard unlikely. (Not my area of experience but offhand this seems rather debatable ?)( maybe appropriate validation / verification data is assumed).
Then again, any idea what the typical meat industry penetration of MDs actually is ? < 50% ?
 rawGround_haz_analysis 05-03-12-1.doc   94.5KB   127 downloads

http://www.meathaccp...raw_ground.html

Rgds / Charles.C

PS - IMHO, the above simply demonstrates that haccp will often not provide a unique solution, especially via generic models. The procedures / meanings of CCP / HACCP are also well-proven to be capable of a variety of interpretations (especially for SQF).
The good part is that auditors also face the same conundrums. IMEX their usual solution is to prepare a set of “standard” expectations based on their own experience/ company’s database / “average” factory responses. When faced with major deviations, objective evidence validating the alternative will be required as per the FSIS document previously attached. Documented cases / references are one popular response.

[MD & X-ray man]

Marshenko,

There is no real limit on the number of items that you can run through a MD so why not run all product through? Is the difficulty the physical handling of the product or the problem with changing the settings?

If it's the latter then here is a possible (although far less than prefect) solution that may help you...

Rather than adjust the MD sensitivity for each different type of product that you run, why not just run with the one setting for everything? Doing this will NOT give you optimal sensitivity for each product. But it will mean that everything will be checked by the MD.

Some other info for you; MD's are affected by a number of different factors, the most likely in your situation being the temperature of the meat and the amount you are putting through at any one time. If you can put all the meat through as frozen (and I mean frozen right through, not just on the outside) then you'll get a much better result. Similarly, if you can put the meat through the MD one piece at a time rather than a box at a time, then you'll get a much better result.

Or purchase a new or newish MD with an automatic learn function and product memories and so that the MD can do all the adjusting for you and you can store all the settings rather than re-learn or adjust everything each time.

Of course if you can get by the auditors without having a MD then that's even easier for you.


Charles,

WRT pathogens, it's not my area of knowledge / expertise so I don't feel that I can comment.



Regarding MD use in the Australian meat industry, I don't have any factual data but, by observation I would say that more than 90% of all consumer meat products i.e. ready to eat, are run through a MD system. Many retailers in Australia will not accept product from a plant that does not use MD as a CCP. If you don't have MD, you can't supply. That's it. Simple.

MD and X-ray man

[moskito]

Hi,

please look into BRC Version 6 - there is a decision tree for using/not using a metal detector acc. to the risk.

Rgds Moskito