7 replies to this topic

[GMO]

The more I read about FSMA the less annoyed with it I get.  When it first came in, my initial response, along with most of the world outside the US was:

 

"you guys literally invented HACCP, why reinvent the wheel?"
 

I'm still (a bit) of that view but am starting to recognise some benefits to FSMA I didn't recognise before, not least because if you read some of the guidance it's getting more towards a retailer standard in the UK or EU than legislation, this stuff is prescriptive in a way that EU and UK legislation never is.  (Which might be why it's in the cross hairs of the current administration and is still in draft but that's off topic.)

 

My point is, I have >20 years of experience in HACCP but there's one aspect of it I'd never really spent a lot of time thinking about.

 

Do you apply your prerequisite controls to your hazard before assessing it for significance or afterwards?

 

In days of yore when I were a wee'un, it was always after but that was before CODEX et al changed their flow diagrams.  The reason being that before that, every significant hazard pretty much became a CCP and you ended up with a ridiculous level of CCPs in a business which could not truly be monitored.

 

So rightly or wrongly, FSMA pushes you back that way but (to my mind) still permits you to have CCPs but the significant hazards which could have been wrongly identified as CCPs in the past would probably now become preventive controls.

 

Is this pushing us backwards or forwards? 

 

I know many HACCP practitioners would argue for the former but I'm minded to think differently.  Yes it was wrong to identify hand washing as a CCP for example, but was it really so hard to control, say, 8 things across a factory rather than 1?  Do our people safety colleagues have this angst in making sure there's only 1 or 2 things to control?

 

It's all got me thinking about how FSMA probably should be impacting plants outside of the US and while I may (probably on these pages) have described the concept of oPRPs as "something for technical people who can't make up their minds"  , now perhaps I'm changing mine and the thoughts of preventive controls managing these significant hazards is no bad thing, if that results in better monitoring and verification of them. 

 

Thoughts?

[Scampi]

I've always done my hazard analysis first following what used to be called FSEP in Canada, which (for certain commodities) required you to assign PRP to hazards following the decision tree, and I will always do it that way

 

A CCP by definition is a point at which, no other controls will lower the risk, seems pretty straight forward to me

 

Relabeling a PRP an oPRP is just semantics

[GMO]

I've always done my hazard analysis first following what used to be called FSEP in Canada, which (for certain commodities) required you to assign PRP to hazards following the decision tree, and I will always do it that way

 

A CCP by definition is a point at which, no other controls will lower the risk, seems pretty straight forward to me

 

Relabeling a PRP an oPRP is just semantics

 

Yes and no.  I think I'm coming round to it after a LONG time of resistance.  I agree in a well implemented plan it is semantics.  But one of the things which is just reality is if you look at most recalls in the world they tend to be related to things which are significant hazards but they're also prerequisite controls and what I don't think all sites are great at is identifying the best verification for their prerequisites or, in doing so, relying on really "late" verification techniques, e.g. consumer complaints.  Or in relying in things like environmental swabbing or product testing they're not recognising that essentially they're detecting a plan that has already failed.

 

I'm just starting to see the point of US legislation and ISO 22000 in this area and I'm torn between "shoot me now" and if I am understanding this correctly "why isn't this explained better and put into one standard?"  

[Scampi]

I firmly believe the allergen recalls are due to sheer ignorance of the importance 

 

Recent recall in Canada because hummus contains PEANUTS and is not declared

 

To which I reply How the hell does that happen?  Is there no batch report?  

Had a hazard analysis been performed PROPERLY, this would have been identified as a possibility and at that point, mitigating steps would have been implemented, I would be willing to bet just about anything that the HA for this particular plant is garbage and no 3rd party audit would ever catch this........they have what, 2 minutes to review the HA?

[GMO]

You're not wrong.  It's probably just me suggesting tweaking around the edges of something that should work if applied properly.

[kfromNE]

One major issue of FSMA - it was put into law before the appendices were complete - so not great direction. Then didn't teach the inspectors how to follow it. At one time we had over 60 Food safety plans. A separate food safety plan for each item we made - even if it was one minor ingredient like adding sliced onions to a product but had the same process. This came from the regional office.  We've shocked quite a few experts over the years. 

I happily can say we have changed directions and got rid of most of our FDA items (various reasons) and also went back down to plans based upon processes so only 2 now. 

 

Being in a USDA/FDA facility - I'd usually take the USDA. They are in the facility every day. They know what happens on a daily basis versus a glimpse. 

 

As for the steps - I have the same thought process for USDA and FDA plans (plus SQF certified) but write then according to the codes (CCP vs PC)  When I can use one program like SOPs - I do it. Then only make separate programs for the USDA and FDA when I have too. If one item, the PRPs, SOPs are written for the stricter code. 

[nwilson]

When I first was implementing a FSMA PC Plan this did change my way of thinking about each process step.  I pivoted to thinking about every possible hazard at that step without the assumption that X, Y, or Z PRP was reducing or controlling the associated hazards.  It made my food safety plans more robust as I had to put the proper controls in place to prevent or reduce to a safe level somewhere in plan.  There was more validation involved and once I integrated HACCP and PC Plan together in one single plan it all came together as something that could be achieved.  Yes, now there are some PRP controlling or reducing hazards it is that very nature of them being a building block in HACCP, however all hazards are identified and are being controlled.  There is more verification to make sure all is operating as intended.  

 

In short, I was not a fan of FSMA either, and when taking my first PCQI course this really opened my eyes to just how many people could not adapt to new things as quickly or just were blatantly reluctant to change.  Now it’s coming around as everyone has been able to digest it albeit in interpretation for most considering all the recalls as of late.  In the last decade I've only seen 4 FDA inspectors across 3 plants I've worked in full time, and not a single one has any issues with my HA, Flow Chart, or Preventive Controls, nor has any customer or GFSI auditor.  This is not to say that each inspector or auditor dug as deep as I would or took the time to do so, but this does make me think that I am at least on the right track and applying a mid-set that helps me sleep as well as I can every night knowing that all things are at their peak performance.  

[GMO]

Thanks nwilson, I first implemented some FSMA type content when the law was first brought in and as I'm UK based it's fair to say I was an early adopter.  I look back now and realised I'd missed the point albeit I wasn't all that far off.  

 

Going back and looking again now, I do think there is a way to combine both HACCP and FSMA and make them more robust as a single entity.  But that's not to say to the earlier point that a well written, well defined plan with great supporting prerequisites wouldn't get there.

 

I think the thing from a UK perspective though is that we have tended not to think about verification as much as we should.  Even in my Campden BRI guideline 42 it feels like an afterthought and we certainly don't automatically think about the verification of prerequisites.  It's all probably there but mostly because the retailers and GFSI demand it but what I've never seen is someone stand up in a daily meeting and say "we've had 3 swab failures in the last week, this is an indicator our HACCP plan is failing".  I have seen people say it's a bad thing or needs some RCA etc but I'm not sure people always make that link if that makes sense?  And when you think about sanitation as an example, we are all so late in the verification methods we choose.  It's always swabbing and product testing when, if you think about it, that shows failure has already occurred.  

 

Although this doesn't mean as I'm suddenly finding some fondness for FSMA that I'm coming around to ATP swabbing which I still maintain is a waste of bloody time.