Jump to content

  • Quick Navigation
Photo

How many times do you run CCP validation tests and why?

Share this

  • You cannot start a new topic
  • Please log in to reply
7 replies to this topic

Jane Jane

    Grade - Active

  • IFSQN Active
  • 1 posts
  • 0 thanks
0
Neutral

  • Germany
    Germany

Posted 29 June 2016 - 07:54 AM

Dear all,

 

I am supposed to validate one of our CCPs. Planning the experimental test realization I was wondering how many times a test run would be needed in order to obtain reliable results. Statistically, I far too high number of runs would be needed, whereas I heard about other sites doing validation test runs only once. Again, other test trials are done three times normally(One positive and one negative, the third one would decide). Since I can't find a guideline concerning technical requirements for CCP validation I would like to ask for your opinion and help.

 

How many times do you run validation tests and why?

 

 

Thank you in advance for your answers,

Kind regards

Jane



Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

  • Earth
    Earth
  • Gender:Male
  • Interests:SF
    TV
    Movies

Posted 29 June 2016 - 09:57 AM

Dear all,
 
I am supposed to validate one of our CCPs. Planning the experimental test realization I was wondering how many times a test run would be needed in order to obtain reliable results. Statistically, I far too high number of runs would be needed, whereas I heard about other sites doing validation test runs only once. Again, other test trials are done three times normally(One positive and one negative, the third one would decide). Since I can't find a guideline concerning technical requirements for CCP validation I would like to ask for your opinion and help.
 
How many times do you run validation tests and why?
 
 
Thank you in advance for your answers,
Kind regards
Jane


Hi Jane,

It likely depends on what kind of CCP you are talking about.

Can you clarify yr product/process/CCP ?

Kind Regards,

 

Charles.C


Sussy

    Grade - AIFSQN

  • IFSQN Associate
  • 39 posts
  • 10 thanks
1
Neutral

  • United States
    United States
  • Gender:Female

Posted 04 July 2016 - 12:11 PM

To validate a CCP, you can either have scientific documentation that determines the levels or criteria it takes for this control be be effective, or you can validate your procedure by getting the correct results consistently, and based on your experience you have validated that the control in place is effective.

 

for example, if your ccp is cooking the product to an internal temp of 145 for 30 seconds to control a specific pathogen, you will either get scientific backup that this pathogen is controlled at that temp, or you will validate your process by cooking it and testing it several times, or over a period of time and if the test results are ok, you have validated this control point. 

 

There is no specific amount of times that you have to perform the test to validate it, but if it is not backed up by sound research and scientific data, then you better make sure it is sufficient to prove the validity.



Scampi

    Fellow

  • IFSQN Fellow
  • 5,444 posts
  • 1507 thanks
1,524
Excellent

  • Canada
    Canada
  • Gender:Not Telling

Posted 04 July 2016 - 03:22 PM

This facility has 3 CCP's, each one is validated with a month of data (20 working days) with each CCP having a minimum of 8 data points/shift, so statistically speaking that is 800 data points at a minimum which should be more that sufficient to validate the CCP along side any scientific studies that compare APPLES TO APPLES


Please stop referring to me as Sir/sirs


Sussy

    Grade - AIFSQN

  • IFSQN Associate
  • 39 posts
  • 10 thanks
1
Neutral

  • United States
    United States
  • Gender:Female

Posted 04 July 2016 - 04:13 PM

Not knowing what particular CCP'S you are referring to, a month of data might not be enough if the circumstances change with the elements.

Just as an example, in the summer it takes 2 days to ferment something and get the LAB to the right levels while in the winter it would be 3-4 days. To establish that, you would need more than 20 days of data.

That is why there is no standard for validating a step. It is very specific to the item, environment, process, procedure, intended consumer etc. Etc.



GMO

    Grade - FIFSQN

  • IFSQN Fellow
  • 2,791 posts
  • 721 thanks
225
Excellent

  • United Kingdom
    United Kingdom

Posted 04 July 2016 - 08:09 PM

Ooh interesting question...  my gut feel from my HACCP training is "you only need to validate once unless you change something" and I will explain why this is both a true and a false statement.

 

Validation is basically "can it work?" so it is a process of checking whether the CCP you have designed can be effective.  In a world called "Ideal world" you would be doing this before you even open the factory and before you process a gram.  Where is your starting point?  Well there are a couple of main ones, literature (e.g. academic papers, retailer codes of practice, legislation, kit manufacturer's validation) and there is testing yourself or by someone you appoint to test for you.  The ideal validation probably has an element of both. 

 

So for example, you are validating a cook process for some chicken.  You would want to use the literature to try and find what target temperature to choose, you would ask your manufacturer to balance the oven and you'd measure it yourself for cold spots so you knew you were targeting worst case scenario, you may then datalog the cook process to ensure the chicken cooks evenly.  You would do some microbiological testing as well and you'd repeat all that a few times and that would be the validation.

 

So the question on repeating it.  Should you?  Well strictly speaking Codex would probably say "not needed" as your verification tests will indicate if something is trending out of control.  But is a good idea?  Hell yes.

 

Is it a good idea to rebalance an oven on a periodic basis?  Yep.  Is it a good idea to crack test a pasteuriser?  Yep.  I suppose though all of these, in a way, fall into prerequisites because they are effectively calibration tests or PPMs which are then supported by verification data.  But people often call it "revalidation".  Does it matter?  Nah.  But I suppose what I'm interpreting your question as is "should I still check some of the parameters I checked as part of my validation?" in which case I would say "yes".  It's not really validation (IMO) but I don't think that matters.

 

This might all seem a tad confusing.  I suppose what I'm trying to say is "it's not massively important what you call it" but also "some of the tests you use as part of your validation can also be used as prerequisite tests and verification".  It all depends on the frequency in my opinion but whatever you call it, repeating some of them is a good thing to do. 

 

That's probably now really unclear...  :shutup:

 

Oh well, I tried my best!  :smarty:



Charles.C

    Grade - FIFSQN

  • IFSQN Moderator
  • 20,542 posts
  • 5662 thanks
1,544
Excellent

  • Earth
    Earth
  • Gender:Male
  • Interests:SF
    TV
    Movies

Posted 04 July 2016 - 08:23 PM

Different Texts / different Product Types / different Food Standards interpret "Validation/Verification" in so many different ways.

 

In the context of the OP a related question is - When is it necessary to re-validate a haccp plan. This is discussed in haccp texts in general terms.

 

I agree with previous posts - for a given CCP I suggest it is necessary when -

 

(a) "Something" changes, either internal/external, so as to significantly affect the risk interpretation of the hazard with respect to the definition of a CCP.

(b) "Something" changes, either internal / external so as to significantly affect the currently implemented Critical Limits.

(c) Nothing is known to have changed for (a.b) but perform a revalidation at least annually anyway.

 

If the OP was referring to the SQF Standard, none of the above may apply. Validation is a "continuous" feature.

If the OP was referring to FSIS, only some of the above probably applies. Validation is, i think, a 2-step feature.


Kind Regards,

 

Charles.C


Neemo

    Grade - Active

  • IFSQN Active
  • 8 posts
  • 2 thanks
0
Neutral

  • South Africa
    South Africa

Posted 19 July 2016 - 11:24 AM

I believe validation is part of monitoring and control of process to drive continuous improvement. Then there's no validation needed

IF

  •  your process(CCP's)  still meets your company specifications
  • you haven't done any upgrades on the process through maintenance or equipment
  • no changes identified through ongoing monitoring




Share this


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users