5 replies to this topic

[AnkushK]

Hello Guys,

Thank you for keeping this discussion platform active all the time !

I need your expertise in designing Pathogen Environmental Monitoring (PEM) Program. We are planning to develop PEM program and have most of the SOPs ready. We have divided our plant in different categories and will be sending samples to outside labs for pathogen testing every week. I want to make a statistical database in which I would like to know that over the time, we are improving our sanitation using some type of index. If there is no index that we can come up with, then please suggest some type of other statistical analysis to evaluate the data. if anything is not clear to you, please feel free to post your questions.

Would highly appreciate your response !

Ankush

[Charles.C]

Dear AnkushK,

Welcome to the forum !

Yr project is a little short on detail Unclear what you mean by "Environment Sanitation" ?

Products ?
Processes ?

Varying / Monitoring / Assessing - What ? / How ?

The most basic index for sanitisers is xD of course.

Rgds / Charles.C

[AnkushK]

Hi Charles,

Thank you for your response !

We are going to start the swabing of product contact surfaces (PCS) and Non-PCS areas. We process various kinds of nuts, so in order to have pathogen free environment, we have developed pathogen environmental program (PEM). Here is the link for very basic PEM program http://www.rqa-europ...re/articles/44.

The plant has been divided into different zones based on the risks and frequencies have been decided based on the risks as well. Frequency of high risk zone is weekly and low risk zone is monthly.

Now my question is, I'll be getting the reports every week from the outside lab. So, I was thinking about some type of sanitation index, say at the beginning of PEM program, I find out more no of positives and later over the time, it decreases. So total positives for one month/total samples for that month tested will give me some type of percentage or index. Next month, when nos of positives will reduce, then index will also reduce.

If there is anything else memebers can suggest, that would be great. Like some type of control charts based on a binomial distribution. I am not really sure, what would be the BEST way to interpret the results ??

Hope I have conveyed my point. If something is still not clear, I would love to elaborate further.

Thanks

Ankush

Dear AnkushK,

Welcome to the forum !

Yr project is a little short on detail Unclear what you mean by "Environment Sanitation" ?

Products ?
Processes ?

Varying / Monitoring / Assessing - What ? / How ?

The most basic index for sanitisers is xD of course.

Rgds / Charles.C

[AnkushK]

Here is the other link: http://ucfoodsafety....files/26483.pdf

[Charles.C]

Dear AnkushK,

Thks for the info and the links.

Unfortunately my browser/protection options (no script) refused to open the first link, gave a very strange partial site display (ie domain just registered...). Maybe there are some odd script usages in the site’s page format. Sorry that cannot access.
I remember seeing the pdf document before here. Amazingly detailed publication although I noted that no actual results are shown. I presume you are focussing on salmonella as far as “Pathogens” go.

I expect that you are aware that this kind of study can be extremely difficult to carry out if the pathogen levels are low (as they should be). I only hv experience of wet environments/doing pathogen analyses for relatively low risk products however I can predict that you are going to need some very good sampling systems / lab backup (as detailed in the pdf document) to try and quantitate pathogens like Salmonella at low levels. This is one reason why the use of indicators like Enterobacteriaceae (EBT) are popular of course although sampling procedures are still likely to need the multiple samples referred in the document if statistical analyses are anticipated, especially in the start-up phase. I am personally more familiar with simple APC type data but EBT data is certainly very popular these days. On the other hand, I appreciate that USA do hv considerable expertise from scanning environments for things like L.mono. build ups. Maybe yr team hv considerable experience in the operational aspects already so you will know all this of course.

My first (obvious? ) guess is that the forward design of the project (such as the control charts you mention) is going to be totally driven by the type/level of initial data obtained, it is always nice to hv in-house back-up for reliability/speed but yr way may give the most sensitive results. Hopefully someone here may have direct experience with yr type of set-up so can make a (daring) prediction.

Rgds / Charles.C

[AnkushK]

Hi Charles,

Thank you again for your response !

I like your ideas and it would be great if someone else wants to give more input in this discussion.

Best Regards,
Ankush