12 replies to this topic

[Plastic Ducky]

Hi all,

 

OK,

So I understand that items rejected by metal detection and x-ray are to be investigated. The purpose of this is to understand exactly what made the rejection take place. I am led to believe the need to know the true nature of the physical contaminate that caused the rejection is toward the value of it's possible relation as a precursor to other problems and in general how you got a physical contaminate in your product.

 

But does this investigation have to begin and be completed prior to shipment? Devil's advocate, if you find a piece of light blue plastic in a product, do you have to open and close the investigation, thereby identifying where this random light blue piece of plastic came from before final release? It feels improbable that you would have to complete the investigation prior to final release because the investigation could take time and be to disruptive to the industry.

 

AND,

What if instead of x-ray, the waste stream to be analyzed was a dud kickout? The dud kick out scenario is different in my opinion because the x-ray investigation one could argue may have implications in the rest of the lot that went undetected as x-ray does not always catch everything. I don't feel that the dud kickout investigation results present the same value as they do not have broader implications to the lot that may have went undetected.

 

So I don't believe either investigation should prevent the finished good that successfully passed all checks and testing from being released.

 

 

Thoughts please...

 

 

[Tony-C]

Hi Plastic Ducky,

 

The need to quarantine and investigate the rejected product would depend on the object found. A significant hazard such as glass or sharp metal would mean I would want a full investigation before considering whether the whole batch should be released. Less so with a piece of light blue plastic bag/liner but more of a concern if brittle plastic.

 

BRCGS Global Standard Food Safety Issue 9 Guidance offer this on false rejects:

Information about false rejects should also be recorded, because this may suggest an error with the detector or indicate that the settings are too sensitive for a particular product and require adjustment.

 

Kind regards,

 

Tony

[paulwill10124]

I would say that if you metal detect every piece of product and one or two are rejected, then you would release the remainder of the product that wasn't rejected (assuming that you are carrying out your start, end and hourly tests).  However, if you are metal detecting one piece an hour for example, then if you have a single reject you would have to quarantine the batch for positive release pending metal detection of every piece in the batch.

 

If you are metal checking every piece of product, you will still want to undertake a destruction test to identify the foreign body.

 

With regard to your example of blue plastic, if this is found in a product during processing then your production team should be able to identify where it has come from (if not, your Technical Team should be able to rapidly identify it).  Once you discover a bit of blue plastic, as in your example, I would instruct the production team and QCs to be hyper vigilant for the remainder of the production run.  I would then only put the product on hold if a large number of pieces of plastic were found because this would be indicative of a serious problem but if it was a one-off then I wouldn't.

[Plastic Ducky]

HI Tony and all,

 

I appreciate your insight.

 

Devil's advocate:

The BRC language indicates that the investigation of false positives has value for the next run of that formulation, but it does not indicate value for the production the observation was noted on, more of a continuous improvement matter.

 

Scenario: LACF product in glass jar.

We investigate an x-ray rejection (x-ray is NOT a CCP). Inside is a piece of glass. This piece obviously was small enough to get through the real CCP (the sieve).

 

No record of a glass break during production, no evidence of a glass break found during sanitation. All operator checks and documentation is correct and complete.

There is your investigation. What is next?

 

1) You could take the entire lot and run it through x-ray again (this is admission that x-ray is imperfect, and duplicating the action still produces an imperfect mitigation)

 

2) You could request destruction of the entire lot. 

 

3) You could record that all points of the investigation into the single piece of glass found in the x-ray kickout produced nothing. And that your investigation is ending with the acknowledgment that without making assumptions you are unable to explain the origin of the piece of glass, and that as there is zero supporting evidence that the issue is wider spread that the single jar that was kicked out that the lot is to be released to commerce.

 

What decision would you make?

[Charles.C]

HI Tony and all,

 

I appreciate your insight.

 

Devil's advocate:

The BRC language indicates that the investigation of false positives has value for the next run of that formulation, but it does not indicate value for the production the observation was noted on, more of a continuous improvement matter.

 

Scenario: LACF product in glass jar.

We investigate an x-ray rejection (x-ray is NOT a CCP). Inside is a piece of glass. This piece obviously was small enough to get through the real CCP (the sieve).

 

No record of a glass break during production, no evidence of a glass break found during sanitation. All operator checks and documentation is correct and complete.

There is your investigation. What is next?

 

1) You could take the entire lot and run it through x-ray again (this is admission that x-ray is imperfect, and duplicating the action still produces an imperfect mitigation)

 

2) You could request destruction of the entire lot. 

 

3) You could record that all points of the investigation into the single piece of glass found in the x-ray kickout produced nothing. And that your investigation is ending with the acknowledgment that without making assumptions you are unable to explain the origin of the piece of glass, and that as there is zero supporting evidence that the issue is wider spread that the single jar that was kicked out that the lot is to be released to commerce.

 

What decision would you make?

Hi PD,

 

3 comments -

 

(1) Yr HACCP Plan IMO looks in serious need of some revalidation/revision regarding hazard analysis/CCPs.

(2) An "appropriate" hazard analysis/CCP(s) would likely answer all yr circumlocutory queries via "Corrective Actions".

(3) Some context is probably needed to make further meaningful opinions. "Devil" in the details ?

[Plastic Ducky]

Hi all,

@Paulwill10124,

 

I feel mostly inline with everything you are stating describing the environment. We are checking every unit. 

I small portion of my question that remains is, do you have to do the destruction testing and finalize the investigation prior to release? I argue in the situation you have described, no. It has no relevance to the lot that has already had a positive release through the x-ray.

 

What does the group think?

[paulwill10124]

 

I small portion of my question that remains is, do you have to do the destruction testing and finalize the investigation prior to release?

 

I would say yes because a destruction test on a single unit isn't going to significantly hold up sending out the product, if at all.  I would want to know the nature of the contaminant and where it is likely to have come from before the product was dispatched because it may indicate a wider problem.  That's just me - I tend towards being cautious.

[Tony-C]

HI Tony and all,

 

I appreciate your insight.

 

Devil's advocate:

The BRC language indicates that the investigation of false positives has value for the next run of that formulation, but it does not indicate value for the production the observation was noted on, more of a continuous improvement matter.

 

Scenario: LACF product in glass jar.

We investigate an x-ray rejection (x-ray is NOT a CCP). Inside is a piece of glass. This piece obviously was small enough to get through the real CCP (the sieve).

 

No record of a glass break during production, no evidence of a glass break found during sanitation. All operator checks and documentation is correct and complete.

There is your investigation. What is next?

 

1) You could take the entire lot and run it through x-ray again (this is admission that x-ray is imperfect, and duplicating the action still produces an imperfect mitigation)

 

2) You could request destruction of the entire lot. 

 

3) You could record that all points of the investigation into the single piece of glass found in the x-ray kickout produced nothing. And that your investigation is ending with the acknowledgment that without making assumptions you are unable to explain the origin of the piece of glass, and that as there is zero supporting evidence that the issue is wider spread that the single jar that was kicked out that the lot is to be released to commerce.

 

What decision would you make?

 

Hi Plastic Ducky,

 

The BRC language indicates that you should act on issues with individual production runs and on adverse trends in number of false rejections.

 

Your scenario with glass in a glass jar:

Assuming it is a long-life product you have time to investigate fully before deciding to release the product. If you cannot identify the source of the glass then it should be sent away for analysis to assist in the investigation, it might be that the typical source can be identified. I have done this previously with glass in product complaints.

 

As you are packing into glass jars in your scenario, it is highly likely that the source is a jar breakage (the analysis mentioned above could confirm if so or not). Jars should be washed or rinsed prior to filling and scanned for defects/foreign objects, is this included in your process?

 

As per Charles’ post, it is quite possible that as you are citing sieving as a CCP, this doesn’t address glass originating in the jar and a revisit of your hazard analysis/CCPs may be advisable.

 

Kind regards,

 

Tony

[Plastic Ducky]

Thank you all for contributing.

 

@Tony,

 

I really like the way you are walking through the scenario. Thank you.

 

The product is in a glass jar. The hypothetical incident if x-ray kick out that contains a piece of glass. The system is a 32 fill head rotary1000 jars a minute production line, so kind of fast.....

 

The receptacle are inverted and blown out with steam prior to fill (so fragments are not present anymore)

The gap being that even if the fill head itself has the sieve (which it does) there remains the possibility of a break happening between fill and closure via capping. There is no way to eliminate this in the 1000 jar a minute system I am looking at. Minimize, yes, limit, yes, eliminate ....no.

 

So if a jar kicked out in x-ray contains a piece of glass sealed inside of it. Which indicates a break in the five feet of automation between the CCP of the sieve in filling and the closure provided via capping the receptacle, then what is the groups next move?

 

The entire lot has been through a positive release, (we don't scan every tenth jar, we scan every single jar with x-ray). The investigation will identify an unknown glass break occurred and the glass is from a jar. So what would you offer next? All jars have been subject to x-ray already, the x-ray is tested with test pouches every two hours, record formed, device working properly. ....

 

So that is the issue, Never been a history of customer complaint out of the line with the x-ray, there is no history driving a lack of faith in this device, everything with a contaminate was caught and everything without a physical contaminate passed through the x-ray. We all know where the glass is from, it's from a jar like the product is packed in.

 

So in front of you you have; 450,000 jars of finished good. One x-ray kick out with a piece of glass inside of it. What's next? 

 

I say the lot is released, the x-ray worked as intended, and the rest of the product has gone through the xray and is to be released. I don't see a line setup that eliminates the 5 feet of automation in between the filler and the capper.

 

Thoughts from the group?

[Charles.C]

Thank you all for contributing.

 

@Tony,

 

I really like the way you are walking through the scenario. Thank you.

 

The product is in a glass jar. The hypothetical incident if x-ray kick out that contains a piece of glass. The system is a 32 fill head rotary1000 jars a minute production line, so kind of fast.....

 

The receptacle are inverted and blown out with steam prior to fill (so fragments are not present anymore)

The gap being that even if the fill head itself has the sieve (which it does) there remains the possibility of a break happening between fill and closure via capping. There is no way to eliminate this in the 1000 jar a minute system I am looking at. Minimize, yes, limit, yes, eliminate ....no.

 

So if a jar kicked out in x-ray contains a piece of glass sealed inside of it. Which indicates a break in the five feet of automation between the CCP of the sieve in filling and the closure provided via capping the receptacle, then what is the groups next move?

 

The entire lot has been through a positive release, (we don't scan every tenth jar, we scan every single jar with x-ray). The investigation will identify an unknown glass break occurred and the glass is from a jar. So what would you offer next? All jars have been subject to x-ray already, the x-ray is tested with test pouches every two hours, record formed, device working properly. ....

 

So that is the issue, Never been a history of customer complaint out of the line with the x-ray, there is no history driving a lack of faith in this device, everything with a contaminate was caught and everything without a physical contaminate passed through the x-ray. We all know where the glass is from, it's from a jar like the product is packed in.

 

So in front of you you have; 450,000 jars of finished good. One x-ray kick out with a piece of glass inside of it. What's next? 

 

I say the lot is released, the x-ray worked as intended, and the rest of the product has gone through the xray and is to be released. I don't see a line setup that eliminates the 5 feet of automation in between the filler and the capper.

 

Thoughts from the group?

Hi PD,

 

(Strictly a hazard analysis needs a flow chart).

 

My next hypothetical moves after an appropriate HA for (I assume) traditional Codex HACCP - 

 

(1) Refer to the Campden Tree for selecting CCPs.

(2) Make the X-ray on end conveyor prior to packaging a CCP and validate it. Implement a typical Corrective Action Procedure as spelled out in many, many MD/CCPs  on this Forum.

(3) Accidents (or false positives) can happen. Re-scanning one,2hr, Production Lot  in 450,000 jars seems not too much of  a big deal (assuming a valid Validation Procedure)?.

 

PS - is the sieve metal construction? Hypothetically > 2 possible hazards ?

[G M]

...

The receptacle are inverted and blown out with steam prior to fill (so fragments are not present anymore)

The gap being that even if the fill head itself has the sieve (which it does) there remains the possibility of a break happening between fill and closure via capping. ...

 

So if a jar kicked out in x-ray contains a piece of glass sealed inside of it. Which indicates a break in the five feet of automation between the CCP of the sieve in filling and the closure provided via capping the receptacle, then what is the groups next move?

...

 

If you acknowledge the possibility of breakage after your sieve-CCP and observe it occurring in your process, it sounds like the x-ray screening of the sealed containers needs to be validated as a CCP to control that physical hazard.

[Plastic Ducky]

All,

 

I want to recognize the x-ray as the CCP in this scenario. But the product is for a special risk group, infants.

 

When I have validated x-rays in the past, we would use a 2-3mm test piece for the validation, and thereafter for all verifications while running. We used 2-3mm as a length in the belief that it was sufficient to meet the choking hazard specification identified by the FDA "7mm-24mm" (CPG 555.425)

 

But that document only states that for special risk groups like infants, the choking hazard range is "less than 7mm".

 

<7 is a range with no bottom. If interpreted literally, it would mean 0-7 which would be impossible. There is not a screen, sieve, x-ray, or metal detector that can be validated to control 0.01mm in a rational volume as needed for processing.

 

So since the FDA fails to give a range with a bottom in this document, how would I validate a traditional control like those noted above? I have found several pediatric references for infants identifying 3mm as the choking hazard max, and that objects <3mm should not be choking hazards. Is the path forward to cite these documents and validate the x-rays using 3mm test pieces?

 

I am grateful for everyone's insight and experience.

 

Sincerely, 

 

The Duck

[G M]

All,

 

I want to recognize the x-ray as the CCP in this scenario. But the product is for a special risk group, infants.

 

When I have validated x-rays in the past, we would use a 2-3mm test piece for the validation, and thereafter for all verifications while running. We used 2-3mm as a length in the belief that it was sufficient to meet the choking hazard specification identified by the FDA "7mm-24mm" (CPG 555.425)

 

But that document only states that for special risk groups like infants, the choking hazard range is "less than 7mm".

 

<7 is a range with no bottom. If interpreted literally, it would mean 0-7 which would be impossible. There is not a screen, sieve, x-ray, or metal detector that can be validated to control 0.01mm in a rational volume as needed for processing.

 

So since the FDA fails to give a range with a bottom in this document, how would I validate a traditional control like those noted above? I have found several pediatric references for infants identifying 3mm as the choking hazard max, and that objects <3mm should not be choking hazards. Is the path forward to cite these documents and validate the x-rays using 3mm test pieces?

 

I am grateful for everyone's insight and experience.

 

Sincerely, 

 

The Duck

 

The US regulatory agencies leave your acceptable threshold up to you to determine based on your risk assessment and the expected consumer.  It sounds like you have some reasonable references for selecting 3mm so yes, I would proceed with the validation with that critical limit.

 

The examples of FM critical limits below 3mm I've encountered had more to do with risk to extrusion equipment than to consumers.