11 replies to this topic

[GMO]

The more time I invest in culture and culture change, the more I think we need to rethink HACCP a bit.

When HACCP came in, it made sense, it was about focusing resource on the points that matter. Fast forward to today and if you look at what causes recalls, they are obvious prerequisites not CCPs.

So this got me thinking. We consider likelihood and severity once the control is applied but what if we brought in a third measure? Likelihood of compliance?

In health and safety they use an assessment of hierarchy of controls called "ERICPD" E is eliminate, R is reduce, I is isolate, C is control, P is PPE and D is discipline. So for example to a health and safety hazard of using a dangerous chemical you might swap that chemical for one which doesn't have (for example) a corrosive rating, that would be a better control than giving PPE (the P in ERICPD) training © and supervision for safe use (D). Likewise in one factory over two years we delisted four out of the five allergens we processed and the last was sulphites. That was a far better control for allergen cleaning than training someone on how to effectively remove allergen residues then checking they actually followed the training (C and D respectively).

So why don't we consider the fallibility of people in HACCP in the same way? I know as leaders this is something we learn through bitter experience but what if we built that into the process?

[AJL]

I love this!! I have done some work on the EHS system (just to get them started). 

When you do the workplace risk assessments, you start with the inherent risk, then the risk with the changes/implementation, then you have a follow up risk assessment in a year or so on. At least thats the template we used. The idea is that the risk level and consequence should both reduce. 

 

With HACCP I find the consequence easy enough to work out, but the risk? Technically you should go after what the risk would be with no control measures, but then everything would just be red  

 

I like the analogy with the EHS  a lot !!

 

When we talk food safety in our plant we always try and minimise risks, as far back as we can, which means we use the EHS principles without even realising it - we like to eliminate! 

We like to try take the human factor out as much as possible

[GMO]

I completely agree and feel the more I learn about health and safety the more I feel like we try and reinvent the wheel sometimes in food safety!

I also think some of this just comes instinctively with experience. But what if we just put in an extra column into HACCP to write what hierarchy of control each control measure is or a likelihood of non compliance? Then I bet it would point you in the direction of putting more attention onto prerequisites...

[Scampi]

Or, companies could just hire good technical staff AND listen to them.......

 

From my extensive experience, this is the single largest hurdle to food safety

 

Bad actors will always be bad actors

 

Of course there is always room for improvement, and rightfully so but if you use risk, and the business is ONLY looking at the risk to profit (as most do) they will NOT eliminate a risk if it comes with an operational cost for that sake alone

[ChristinaK]

With HACCP I find the consequence easy enough to work out, but the risk? Technically you should go after what the risk would be with no control measures, but then everything would just be red

 

I believe that's what FSMA in the US has us do (HARPC)...analyze the likelihood of the hazard in the absence of process controls and prerequisites.

 

Isn't the person element one of the reasons GFSI came out with Food Safety & Quality Culture?    Although I do agree that sometimes it is best to remove the possibility of an employee being able to "screw it up" so to speak.

[GMO]

Yes absolutely the person bit is the food safety culture and that will always be a factor. What I'm trying to get at I suppose is more around the ergonomics of the design. I've how close will it be between the work as it was designed vs what happens in reality. Of course bags of that is culture but, for example, if your control measure for a microbiological hazard is effective cleaning and you could get a utensil washer Vs manual cleaning... While utensil washers are not perfect they will, in general, give reproducible results whereas an operator may not. They may cut corners, not dilute the chemical correctly etc Vs the dosing system on a washer...

But as Scampi rightly points out, investment happens when there is a return on it. Not to design out food safety risk (or only if that's a happy coincidental outcome). But imagine if we had similar laws in the UK for example to H&S where there has to be a "reasonably practicable" level of risk avoidance? Sure H&S funding isn't perfect either but I've had funding turned down for H&S and quality improvements which would cost peanuts in comparison with the profitability of a company.

I just think we're slightly missing something here that our H&S colleagues consider when it comes to compliance. Why are we ignoring that human factor in our risk assessment stage? Before we get into engagement, which I agree is bloody huge, why not spent a small amount of extra time on HACCP making it better and properly identifying where compliance is likely to be poor then at least targeting our verification activities in a more playful way?

[GMO]

Planful... Not playful... But now I'm chuckling to myself about playful verification... 😆

[G M]

I suppose the alternative is just recognizing that risk is chronically underestimated. 

 

The core of the HACCP plan is risk assessment.  If the proposed conclusion is that the outcome of reduced and controlled hazards is being oversold by the HACCP process -- then wouldn't those be properly controlled if the risks were accurately represented and suitable controls put in place?  That is fundamentally what those environmental health and safety programs are doing, acknowledging the risk of inconsistent compliance, and putting in place more controls to manage that additional risk.

[GMO]

Yes you're right GM.  I suspect the HACCP purists will say that you should be accounting for compliance into "likelihood" but there isn't a prompt in any plan I've looked at which directly asks this question.  

 

In the meantime I'll march forward with my flag of "don't forget people may say they do things but often don't" flag of non compliance.  

At a base level, any control which is based on a person should have something they're trained against and can be audited against.  BUT if that is a person we should also be putting a CI plan into our HACCP process to improve the hierarchy of control should we not?  We would if this was health and safety.  If M&S have their way M&S suppliers will probably all have these plans for quality but not necessarily for food safety.  Crazy?  Yep.

 

So fast forward and think about what this could look like.  For example, various retailers and group technical teams have decided in the UK that online barcode or 2D barcode verification is a standard requirement because of the level of mispacks which have occurred over the years and because of the allergen risks to a consumer if that occurred.

BUT what if you had a lens on your HACCP plan on that prerequisite control which we know as an industry has caused vast numbers of recalls (likelihood high) and could kill someone (severity high) but because a manual check is considered a control measure, it took standard changes to require manufacturers to install more control and push it up that ERICPD hierarchy.  BUT if you had a site really thinking it through, they may plan differently.  Even get a bit more oomf behind delisting allergens which do not have functional properties in the product for example.  

I just think we're missing a trick...

[GMO]

Reflecting on this again today.

Recently we've had an incident where we had failure of a heat process.  If that failure occurs the staff are meant to manually scrap the product and put the machine into a sterilisation cycle.  

 

However, there is a way to bypass this.

Guess what happened?

 

Another issue.  Same place.  There was a CI project to save the small amount of ingredient left in an IBC / bulk container.  So to do this they reprogrammed the pump to continue operating even if dry.  However, the automated system registered delivery using the pump.  So dry pumping would count as ingredient addition.  To counter this, they put in a manual weight check.

 

Guess what happened?

 

I think all of us who are a bit long in the tooth like me can see the obvious risk inherent in both of those processes.  And yes, a strong quality and food safety culture will help mitigate people cutting corners.  But also people are people.  And if you can design out a food safety or quality risk at pretty low cost (for example in the second case, run the automated system off the load cell rather than the pump) then just f---ing do it, right?

[G M]

... then just f---ing do it, right?

 

But I don't wanna have to remember/learn/do new things. 

[GMO]

I am mulling this over repeatedly at the moment.  I do think HACCP is valuable but we put HACCP mostly into the hands of (no offence), people like us.  People who like science and techy stuff and not necessarily people who are good at understanding others, what motivates them, what motivates non compliance and the (frankly, much as I love my operations colleagues) the bunch of cowboys running the plant.  

 

So... stay with me here.  What if we adapted HACCP and had something like:

 

Severity, likelihood, control measure, hierarchy of control measure (we'd need to come up with a different acronym than ERICPD but imagine using something similar), then for everything which was more than a low hazard but controlled by discipline, we put focus into improving those areas.   That might be a combination of investment, auditing etc.