Is PDCA the correct methodology for ISO22000?

Hello eveyone,

I've just been looking over my training course notes for my ISO22000 Lead auditor training and have come up with a bit of a tricky situation that hopefully all of you food safety practicioners out there could help me resolve.

During the training we spent a lot of time discussing the Plan Do Check Act process integral to Quality Management

at the time of the training ISO22004 was not available but having looked at the flow chart on page vi (Simon can you let me know if I can post it here?) and spent more time looking at the validation requirement in ISO22000 the PDCA approach does not seem to be appropriate (and in fact contradicts the validation requirement of ISO22000).

As I interpret the standards I think that there is a different methodology required, the Plan Check Do Check Act approach (for modesty's sake lets call it the Gibb principle ).

This would explain why all the companies I have seen so far have not understood the concept of validating of control measure prior to implementation - If they follow the PDCA approach they are not expecting to validate the control measures until after their implementation.

Does anyone have a view on this, I'm open to having my new principle shot down in flames.

James

James,

I like the gibb principle - its certainly in the right direction and better than PDCA which is really for QMS. You are right about the need to validate the design of your control documentations to meet the purpose of the intended objectives. i.e. pre-implementation validation

Once implemented under draft, do the results or strength of control serves to achieve the desirable objectives intended. Document approved and issued post-Implementation validation

IMO, PDCA is in fact crap as far as ISO 22000 implementation is concerned. The cycle for continuous improvements is much more demanding as we are referring to a dynamic food processing system.....which will lead to continous subsequent system and process validation (remember.....combined control measures principles )

Hello eveyone,
Simon can you let me know if I can post it here?

In the interests of furthering debate I don't see a problem - please go ahead James.

I've got a heavy cold so I can't think to much just now. All I know is with PDCA you check (validate) after you have acted (implemented) so the Gibb theory may well be the WTG for ISO 22000.

Simon

Simon,

I think you must have a computer virus as I'm now also ill!

attached is the flow chart from 22004

Pages_from_ISO22004.pdf   64.92KB   267 downloads

Hopefully this will stimulate some further debate?

James

Although it looks good, IMO the continuous improvement cycle does not reflect the requirements demanded by the standard particularly in the demand for verification and validation aspects. Staging a document to achieve control of an intended objective requires to be designed for effective result arising from the data recording. Does it achieve this? If not, we would need to redesign.....or in more severe option, modify the process and afterall all these, revalidation is definitely a must before moving forward to effective verification and leading to currency and validity of the system.

40 downloads but not much debate; does this signify people don't understand the model in 22k.

James perhaps you should explain your model in some more detail.

Regards,
Simon

40 downloads but not much debate; does this signify people don't understand the model in 22k.

James perhaps you should explain your model in some more detail.

Regards,
Simon

Simon,

quite simply ISO22000 requires a pre-implementation validation of control measures.

call me old fashioned but for me pre-implementation must occur before the "DO" otherwise it is post-implementation.

Validation is a "Check"

So surely there must be a "Check" before the "DO"?

In the flow chart Validation is clearly shown before implementation, I know Charles is not too enamoured with the flow chart but it does show the processes in the correct sequence as specified by the standard.

The big problem with validation is that people are confusing "validation of control measures" with "Product validation" from Quality systems and the 2 are different

James

Simon,

quite simply ISO22000 requires a pre-implementation validation of control measures.

call me old fashioned but for me pre-implementation must occur before the "DO" otherwise it is post-implementation.

Validation is a "Check"

So surely there must be a "Check" before the "DO"?

In the flow chart Validation is clearly shown before implementation, I know Charles is not too enamoured with the flow chart but it does show the processes in the correct sequence as specified by the standard.

The big problem with validation is that people are confusing "validation of control measures" with "Product validation" from Quality systems and the 2 are different

James

Thanks for the clarification James; to make it really clear can you give an example of a pre-implementation validation of a control measure.

Ta,

Simon

Simon,

probably the best example is the metal detection debate currently raging in another topic http://www.saferpak....owtopic=317&hl=

another example would be:

I need to use a combination of hot water, cold water, detergent & disinfectant for hand washing.

will cold water and detergent work sufficiently well to ensure clean hands (the cheapest option)

or do I have to use hot water and disinfectant? (the most expensive)

The validation method will depend upon preference, if the disinfectant manufacturer garuntees 99.9% bacterial kill with cold water is this sufficient for your needs?

initially I would carry out a series of swab tests to satisfy myself the effectiveness of the different processes depending upon the results I would then select the best (ie cheapest method that gives the desired results) method. (pre-implementation validation)

typically I would carry out occasional swab tests to confirm the control measure is still effective (verification)

and if a cheaper disinfectant came onto the market (any change therein..)I would carry out swab tests to check the effectivness prior to using the new disinfectant.

Does anyone else have other examples?

James

initially I would carry out a series of swab tests to satisfy myself the effectiveness of the different processes depending upon the results I would then select the best (ie cheapest method that gives the desired results) method. (pre-implementation validation)

James,

If the effectiveness of the disinfectant was not known or advised on the label as per scientific validation caried out by the manufacturer in the first place, I would not even bother to verify it as there is no benchmark to validate. Simply, it would not withstand the approved supplier criteria.

IMO preimplementation validation should be conducted on the effectiveness of personnel training on the washing techniques, amount of dilution ratio used including the where, when and why etc for effective personnel hygiene rather than conducting a pre-implementation validation on the detergent's bactericidal strength which is a prerequisite anyway. Frankly, in this instance, I would only consider risk assessment base on Control of Suppliers Specifications Program while a test if needed is only a preimplementation verification rather than a validation as it is merely a confirmation that the supplier is indeed telling the truth against whats been claimed.

I guess under ISO 22K the preimplementation of V & V covers a much wider scope against different purposes under the ever important combined control measures approach as demanded by the standard.

Charles,

I agree with you, I was trying to give a simple example, I guess nothing is simple with Food safety.

BTW I was thinking more the hot water vs cold water issue, most guidelines say use hot water but many washing facilities only use cold water.

James

Charles,
BTW I was thinking more the hot water vs cold water issue, most guidelines say use hot water but many washing facilities only use cold water.

How warm does water have to be before you can call it hot?

How warm does water have to be before you can call it hot?

Bingo!

That's why we have to validate.

James

How warm does water have to be before you can call it hot? doh.gif

Simon
I find it odd that we should take guidelines from the standard so seriously as afterall they are guidelines. Most if not all manufacturers would provide the necessary application direction anyway.....whether its to be used with hot or cold water (or worse something in-between )

This is where ISO 22K is so pragmatic. It does NOT tell you what to do or should do. There are NO guidelines but it merely suggests that you "Do what you like" but as long as you can proof to us that it is suitable for your organization (complexity or industry type etc) and has been verified to be effective in controlling your intended objective(s) then that should be fine.

I guess nothing is simple with Food safety.

James,
Thats probably the reason why we find dealing with food safety so exciting because of its dynamic nature. Different angle of opinions can prevail and really, there is no single correct answer to every thing in FSMS. All that is really needed is proof of control evidence.............and that, means everyone may be corect in their own rights!

I had resisted joining this thread since I have not read ISO22000 however since Simon has expanded the scope -
I would think cost is an unavoidable issue in the choices discusssed above ?
Unless you work for the 'big boys' I suspect many people lack the ststistical background to create validation studies for imprecise parameters such as microbiological measurements although it is usually rapidly evident if something is simply useless (or being wrongly implemented), eg consistently 'astronomical' swab counts on hands / gloves.
It is my understanding that PDCA is a recommended interpretive approach to ISO9000, not a specific mandatory thing to be audited / passed on. I scanned over it via it's link with BRC food, noted it looked complicated and was grateful for not having to work directly with ISO9000. However now I see it transferred into a HACCP format it seems more sensible since HACCP already constitutes a logical series of steps (yes?). In this respect I find the flowchart attractive however the bottom dotted loop looks almost like an afterthought and CharlesChew is surely correct regarding revalidation.
Regarding PCDCA, perhaps one could redefine 'Planning' and include the 'C' in the case of validation though the Gibb's cycle has a nice ring to it.
Rgds Charles.C

Dear Charles C.

Revisiting some of the older threads (which are still very useful), you have raised useful comments on the PDCA approach and their cross functional activities along the way for verification and validation activities.

(After having taken a few companies to ISO 22K Certification), from my audit experience, it is necessary and appropriate to verify and validate your systems during the following stages: design stage, implementation stage and post implementation stage. This way you would be able to demonstrate whether your present process environment is producing food with the appropriate level of protection.

14-year old thread.

It is very interesting topic to be debated..

 

In my opinion validation / verification also to be considered in PDCA cycle, as explained in above example of control measure validation. As you can see in ISO 22000-2018 version, there are 2 PDCA cycles operating at the same time. 

First PDCA cycle -

PLAN - Context of Organization, Leadership, planning and Support.

DO- Operation

CHECK- Performance evaluation

ACT- Improvement

 

Again Under clause 8.0- Operation contains second PDCA cycle in which validation of control measure comes under "PLAN"

PLAN- PRPs, Traceability,Emergency preparedness, Hazard analysis, validation of control measure,Hazard control plan & Verification planning.

DO-Implementation of PLAN, Control of monitoring & measuring, control of product & process NCs,

CHECK-Verification activities, Analysis of results of verification activities

ACT- Updating of preliminary information & documents specifying PRPs & hazard control plan.

It is very interesting topic to be debated..

 

In my opinion validation / verification also to be considered in PDCA cycle, as explained in above example of control measure validation. As you can see in ISO 22000-2018 version, there are 2 PDCA cycles operating at the same time. 

First PDCA cycle -

PLAN - Context of Organization, Leadership, planning and Support.

DO- Operation

CHECK- Performance evaluation

ACT- Improvement

 

Again Under clause 8.0- Operation contains second PDCA cycle in which validation of control measure comes under "PLAN"

PLAN- PRPs, Traceability,Emergency preparedness, Hazard analysis, validation of control measure,Hazard control plan & Verification planning.

DO-Implementation of PLAN, Control of monitoring & measuring, control of product & process NCs,

CHECK-Verification activities, Analysis of results of verification activities

ACT- Updating of preliminary information & documents specifying PRPs & hazard control plan.

^^^^^ No offence intended and I wish I could agree with you more than 14 years ago but times have not changed for myself. I still find the PDCA topic one cause for avoiding use of ISO9001/22000.

 

Nonetheless, there are obviously millions of ISO enthusiasts who disagree with me. :smile:

14-year old thread.

For information attached to this topic is a poll, when someone votes in a poll it bumps the topic to the top of the recent topics list. It may be 14 years old, but still can be debated. In fact it still is a relevant topic as can be seen in our recent webinar recording at the link below.
 

An Overview of FSSC 22000 V. 5
The FSSC Foundation published version 5 of their standard in May 2019. This new version of the scheme has been issued to primarily include the new ISO22000:2018 as well as to simplify and clarify some points from the previous version. A key change with the new ISO 22000:2018 scheme is the inclusion of 2 Plan-Do-Check-Act (PDCA) cycles within the FSSC 22000 scheme. This presentation will look at how these 2 PDCA Cycles are involved in looking at the organizational risk and operational risk in both a positive and negative manner.

https://www.ifsqn.co...-fssc-22000-v-5