Jump to content

  • Quick Navigation
Photo

How to formulate and document a risk assessment

Share this

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

D-D

    Grade - MIFSQN

  • IFSQN Member
  • 239 posts
  • 55 thanks
5
Neutral

  • Canada
    Canada
  • Gender:Male

Posted 20 October 2010 - 07:41 AM

Just looking at some of the standard BRC risk assessment requirements:

7.3.4 - Hand washing - if our policy is to wash hands before starting work or after breaks/smoking/toilet/nose blowing, what is there to risk assess?
5.5.1/5.7 - Product release - if QC test every batch (flavouring products) against full spec, what is there to risk assess?

It would also be easy to write into some of these assessments things like "we have been doing it this way for nearly 50 years without incident so current practices are acceptable". Talking to customer auditors it seems this is not good enough but it is sometimes difficult to understand what is required.

I have a nice template to use now but struggle to fit it to what we are doing!



Charles.C

    Grade - FIFSQN

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

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

Posted 20 October 2010 - 11:51 AM

Dear D-D,

The fact is that posters here seem to have achieved auditor satisfaction in a variety of ways ranging from a formal HACCP type layout to a ‘simpler” explanatory style. Not sure what kind of template you have however the 2 links below contain a variety of general IT / specific BRC targeted examples and (cross-linked) discussions on (BRC) expectations. The condensed chart by baron is perhaps the widest in (BRC) scope but seemed (to me) to demand further expansion of a specific entry, eg handwashing (or perhaps not ??)

http://www.ifsqn.com...dpost__p__24232

http://www.ifsqn.com...dpost__p__32862

IMHO, some variables are relatively amenable to quantitative treatment (if desired) whereas others are more qualitative in nature. Just like HACCP and thereby offering equal subjectivity. :whistle:

Maybe you could post yr template to add to the collection and inspire others ? :smile:

Rgds / Charles.C


Kind Regards,

 

Charles.C


D-D

    Grade - MIFSQN

  • IFSQN Member
  • 239 posts
  • 55 thanks
5
Neutral

  • Canada
    Canada
  • Gender:Male

Posted 20 October 2010 - 04:02 PM

Maybe you could post yr template to add to the collection and inspire others ? :smile:


Hello Charles,

Template attached although it will probably be of limited value as you will see all the resources you pointed to have all been well plifered! I have attempted to fill it out for work wear e.g. risk of contamination by wearing it in the canteen or outside etc and input from a colleague is reasonably favourable...

Thanks,

D-D

Attached File  D-D Risk Assessment Template.doc   86KB   1629 downloads


Thanked by 3 Members:

Charles.C

    Grade - FIFSQN

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

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

Posted 20 October 2010 - 09:49 PM

Dear D-D,

Thks for attachment. As you say, well-pilfered but result looks quite functional. :clap: I agree with you that it is not so easy to produce a (simple) totally generic form for all types of situations.

I tried it out for handcleaning and the result is attached. Not a criticism but I am used to slightly different format / risk matrix so I took the liberty of temporarily adjusting a few things. (I tried to mark the changes in italics / bold format and my risk matrix is added at the end of attachment).

Attached File  testing3 (handcleaning) - D-D Risk Assessment Template.doc   124KB   1256 downloads

Rgds / Charles.C

added - Not sure if column / item BBBB is overkill, if included, should definitely be genericalised as much as possible similar to (borrowed) example I posted in Gemmas thread. In this specific case any verification micro results can be overlapped to the routine SSOP program.

added2 - with regard to yr query as to "what risk to assess" I agree with you again that there may be a degree of intuition / subjectivity involved. Where the hazard is spelled out such as "accident to worker", the scope is wide but the primary objective is well defined, on the other hand , specifying a control procedure (possibly multi-function) implies a reverse thought process which may not be explicit, eg pest control is not targeted at pests, at least from a scientific viewpoint :smile: , but at their disease potential. Furthermore their presence is linked to other factors, free food supply, well-concealed, etc. I suggest you hv a look at my post http://www.ifsqn.com...post__p__24256. Other people may have a more precise decision logic??


Kind Regards,

 

Charles.C


Thanked by 3 Members:

D-D

    Grade - MIFSQN

  • IFSQN Member
  • 239 posts
  • 55 thanks
5
Neutral

  • Canada
    Canada
  • Gender:Male

Posted 21 October 2010 - 02:06 PM

I like it Charles; thank you for your input. With a simple bit of re-phrasing and the example it makes it a lot clearer.

Are you saying I could ditch the "Verify Controls are effective" columns (Review method & Date completed)? That would be good as I could use the space to put names agains the actions but I will keep them in if so adivsed.

Thanks again.



Charles.C

    Grade - FIFSQN

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

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

Posted 21 October 2010 - 07:35 PM

Dear D-D,

Well, from other reported experiences, it appears that some people have found it auditorily un-necessary to even maintain a documented risk analysis. (Note that such a requirement does not routinely appear in all individual cases in the standard). Furthermore, from the list posted by Tony, approx 27 items are risk – analysis related (!), a substantial amount of work. For such reasons, it seems practical to perhaps include a lower, generic paragraph headed, for example, “verification” to each documented risk analysis stating something like – the above implemented actions / procedures / risk bases will be re-examined for correctness in case of any related instances of loss of control, changed conditions or, at a minimum, on a yearly basis. This is analogous to methodology used for the HACCP program. This allows use of a condensed “verification” document listing risk analyses vs audit check date.

Might be easier to do an Excel based version allowing several BRC clauses to be put together, or perhaps using individual sheets. Question of Word / Excel expertise ?.

Interested to hear other people's approaches / experiences :whistle:

Rgds / Charles.C

added -

After little more musing, a few additional thoughts –

the reverse thought process I mentioned can perhaps be summarised –

1. controlling what primary type of hazard ? > safety / legality / quality? > specific?, eg BPC, weight, OLQ > Hazard Column.
2. hazard most likely caused by what activity ? > data / conceptual > Activity column (this approach is clearly pre-selective but so-what? It's undeniably a risk assesssment :smile: ).

Referring Tony's BRC "want" list, I guess the majority of items are "top level" safety wise and their risk-related aspects / explicitly-implicitly associated inspection frequencies will probably hv already been considered within existing programs.
Some are seemingly less incident-prone, eg jewelry, wood. De-emphasise ?

The comments regarding verification will clearly require some x-referencing to related procedures which probably already exist.

Other RA routes are obviously possible, eg specifically delineating the risk-assessment aspect within the actual appearance of the item in one's manuals and then compiling a master document with x-references. I encountered a similar quandary with "Validation" for HACCP but ultimately chose to append a specific chapter to the HACCP program. Question of preferences again.

NB - OLQ = organoleptic quality :oops:


Kind Regards,

 

Charles.C


Thanked by 1 Member:

D-D

    Grade - MIFSQN

  • IFSQN Member
  • 239 posts
  • 55 thanks
5
Neutral

  • Canada
    Canada
  • Gender:Male

Posted 22 October 2010 - 07:35 AM

Dear Charles,

Thanks again for your help; this continues to be a very useful thread for me and hopefully others too.

I had already considered using Excel to free myself of the limitations of an A4 page. I have been resisting it as I spent so much time making my Word table look nice but I think it is the best way to go (I have seen some very complex Excel based RA's from customer auditors who were kind enough to share their approach).

I had also thought of putting the 'verification' into a box at the bottom where it can be confirmed and signed off rather than doing it for each and every action. As you suggest, an alternative could be similar to HACCP reveiw and report an annual RA review through a minuted meeting where the effectiveness has been discussed and signed off.

D-D

Added:
Below is the image of the top of my new Excel based RA table; hopefully it is clear to see. QSL = (Product) Quality/Safety/Legality, CPMA = Chemical/Physical/Microbiological/Allergen (Hazard). It's much better like this, I can use a calculation to multiply out the Risk factor and with a bit of nifty conditional formatting (my favourite!) make the cell go red/orange/yellow depending on the rating.

Attached Files


Edited by D-D, 22 October 2010 - 09:00 AM.


Charles.C

    Grade - FIFSQN

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

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

Posted 22 October 2010 - 11:46 AM

Dear D-D,

Yes, certainly more numerical / sizing options in Excel. I have had the same experience as you, reluctance to relinquish dozens of tables laboriously but elegantly (eventually ) constructed in Word. Even in landscape / legal size, 20 column tables / complex formatting are a challenge in minaturisation. :smile:

One passing thought, yr current format could perhaps be expanded into an overall structure similar to that posted by baron which (despite my earlier reservations) is the only cumulated framework I have yet seen here (I think). Some of the “blank” squares in his chart well illustrate the conceptual difficulties in responding to some of the (nutty) BRC expectations in a condensed form. Some programs, eg Supplier Approval, are essentially already RA documents for many people so should really only need a x-reference IMO.
Just a question of personal preferences. :smile:

Rgds / Charles.C


Kind Regards,

 

Charles.C


Thanked by 1 Member:

D-D

    Grade - MIFSQN

  • IFSQN Member
  • 239 posts
  • 55 thanks
5
Neutral

  • Canada
    Canada
  • Gender:Male

Posted 22 October 2010 - 01:14 PM

Dear Charles,

Currently I am using the baron spreadsheet as a prompt. It is very useful but I wonder if one could be accused of being a bit 'brief' when covering all 27 or so RAs on one page (auditor dependent maybe...?). Some parts I don't understand either e.g. facial hair as an allergen hazard.

In any case I am feeling good about my shiney Excel chart and have already used it for section 7.5. I plan to put all the BRC RAs into one workbook with a different one on each worksheet (or grouped as appropriate). By copying and pasting the format to start a new one it is going to save a lot of time. Glad I abandoned Word; it didn't take too long to change over anyway.

By the way: If you compare the vanilla BRC clauses with the Interpretation Guideline book, "risk assessment" crops up more than the standard 27 that Tony gave us e.g. 7.3.1.

Latest blanko attached in case anyone wants to develop it further.

Thanks again and good weekend,

D-D

Attached Files


Edited by D-D, 22 October 2010 - 01:40 PM.


Thanked by 2 Members:

Charles.C

    Grade - FIFSQN

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

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

Posted 23 October 2010 - 12:15 AM

Dear D-D,

Looks OK. Nice auditor - restful colour. :rolleyes:

One item which may be required (somewhere) is the ubiquitous revision statement(s), eg “revised on”

By copying and pasting the format to start a new one it is going to save a lot of time

I hv never routinely used the function but (if dividing by worksheet) I think you can directly save the chosen layout as a defined Excel template (eg tp1.xlt) > close. Then - (right click) on “Sheet X” at bottom of default sheet of any book > select insert on drop menu > select tp1 on next drop menu > OK. I just tried it on Excel 2003 and seems to work.(the sheets can be moved around [drag-drop sheet No.] at will of course) I daresay there is a more auto-requesting method for the Excel regular. :smile:

Rgds / Charles.C

Kind Regards,

 

Charles.C


D-D

    Grade - MIFSQN

  • IFSQN Member
  • 239 posts
  • 55 thanks
5
Neutral

  • Canada
    Canada
  • Gender:Male

Posted 07 June 2011 - 02:00 PM

Just coming back to one of the favourite subjects...

I have been using the risk assessment matrix that was discussed above but something puzzles me:

There are 4 colour coded Risk severity areas on the matrix: green, yellow, orange, red. However, if the Likelhood of a hazard occurring is reduced to 1 (i.e. Rare), why would the Risk on that level of Likelihood ever be elevated from green to yellow even if the Consequence is a 4 or a 5? If the Likelihood is the lowest score, surely the whole row should be green and not require much further consideration irrespective of the Consequence rating?

Attached is a 3 colour matrix I got from a recent Health & Safety risk assessment course, which does exactly this. Is this not better to use...? Basically it is combining all the green and yellow into just green.

Attached Files



Charles.C

    Grade - FIFSQN

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

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

Posted 07 June 2011 - 10:10 PM

Dear D-D,

Yr observations illustrate the handling of the subjectivity factor in risk assessment, ie how to compare the relative significance of the various (likelihood / severity) combinations which may occur.

Some users opt for a “severity” prioritised logic in the sense of their preferred outcome decision (ie rating) so that, for example, “catastrophic”/”rare” is classed as moderate rather than low (sometimes the result may even be > moderate). Other users follow the opposite logic as shown in yr second example. Similar decisions are involved across the table. The size of the matrix also influences the options available.

The (risk) philosophy often recommended in food HACCP plan development books is that likelihood of occurrence should (effectively) take preference over severity, for example, the “justification” column in the popular table format often used in USA implicitly requires that the likelihood of occurrence be “reasonably likely” or more. But maybe not for astronauts ? :smile:

A lot of related decision theory originated from assessing the risks of operating nuclear reactors. The potential effect of an incident could be calamitous which perhaps dictates an ultra-conservative approach but ……

Presumably the chosen methodology should be validatable (somewhere). :smile:

Rgds / Charles.C


Kind Regards,

 

Charles.C


D-D

    Grade - MIFSQN

  • IFSQN Member
  • 239 posts
  • 55 thanks
5
Neutral

  • Canada
    Canada
  • Gender:Male

Posted 08 June 2011 - 08:33 AM

Thank you for the comments Charles. Perhaps I will stick with the 4 colours you showed me originally.
Something else a little more basic: When performing the initial LxC, should the values used consider controls already in place or should they be theoretical values based on no controls (which would therefore make them high) that are then reduced by the existing / future controls...?



Charles.C

    Grade - FIFSQN

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

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

Posted 08 June 2011 - 08:17 PM

Dear D-D,

Yr question is another popular one which is answered in various ways in different texts. I recently came across the university (teaching) material attached below (2009) which contains IMO some quite nice in-depth looks at key aspects of HACCP analyses. Their response to yr query is –

For all listed hazards, the food industry should investigate the chance or probability that the mentioned hazard is occurring and if it occurs, what the effect on public health may be. In other words the risk depends on the probability and the effect: risk = probability × effect.
(added - many books replace “probability” with “likelihood of occurrence”)

What is the possibility of a negative effect, when the control measure does not work well?

Probability = the risk that the hazard occurs in the end product when the control measures are not present or are failing and taken into consideration the implemented PRP measures
(added – the underline is original)

“End product” implies that all next steps in the production process are taken into consideration so that correction in a next step in the production process is possible .
(added – for “completed” end-products, many [I think most] books also include treatment [assumed properly carried out] by the consumer, eg cooking ) (this methodology may create a conclusion at odds with legislatory requirements in some cases).

Effect = the effect of this hazard on the health of the consumer


(You can see yet another presentation style of risk matrix on same page (117) as above with a possible action rule set given on pg 125)

Attached File  QM Systems in food 2009.pdf   789.2KB   776 downloads
Rgds / Charles.C

Kind Regards,

 

Charles.C


Thanked by 1 Member:


Share this

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users