With an audit coming up very soon I have been drawing up a micro testing sampling plan. The products we are sending away will be uncooked ready to eat prepared vegetables. I spoke to our food safety consultant about corrective actions for a failed micro test and he suggested the documented action would be to just keep sending off further samples until a reading within the specified limits is reached. This doesn't sound correct or even ethical to me as if there is a problem with the process it feels like something should be changed rather than just keep on trying until it passes. Can anyone lend any further advice?
Sack the consultant?! Seriously? I think as you've identified yourself that's a pretty rubbish response.
Your test results will show one of two things; an indicator, e.g. TVC, enteros which indicates poor hygiene or a pathogen. I would split things into two aspects; corrective action (immediate) and preventive action (which follows an in depth root cause analysis.)
So, for example, you have a product in the marketplace with a pathogen, your immediate correction might be to recall it or hold it although with shelf life this might not be possible. Next root cause analysis. What went wrong? Check through the processing, check the records but also what is happening now. Sample components. Check machine hygiene. Swab the environment. Go deep into machines. This micro result is a verification (or lack of) of your HACCP plan so, essentially part of your HACCP plan has failed, whether that's a prerequisite or a CCP. Your response and the amount of investigation you do will be commensurate with risk. So, for example you might not choose to investigate every entero failure but you might react to a trend.
A great thing I did many years ago (not in fresh produce though) was to list my products in % failure order, take the top 20% and work out what the common factors were. After doing this, I identified one machine which steamed vegetables. I stood watching the team members using this machine on days. All fine, temperatures fine. Then I watched on nights and realised they were cutting corners resulting in an ineffective cook. Simples. Halved my out of specs as a result.
Once you've decided what you do intend to do (e.g. react to trends on some indicators, react to incidents on pathogens) put it in a procedure as well.