16 replies to this topic

[Chac]

Hello everyone!

From time to time we send a sample our products to an external laboratory for analyzing them. Now we've got the results and they found more microorganism than our internal laboratory. The've got the same sample that we have used, they used the same methods.

How is it possible that they have founded more mo than we have?

[Charles.C]

Dear Chac,

Nice to hear from you !

Unfortunately the amount of info. you provide would require a crystal ball ? Much more please.

Rgds / Charles.C

[Cathy]

There are many, many factors that could be involved. Did you analyze your sample on the same day you took the sample? Did the sample sent to the outside lab take some time to get there? This can allow some growth even if the sample was cold. Did both labs use the same method ?

Bacteria are rarely distributed evenly in any sample. Unless the variation is very large, it is normal to see variation within the same sample. Microbiology isn't like chemistry !

[cazyncymru]

There are many, many factors that could be involved. Did you analyze your sample on the same day you took the sample? Did the sample sent to the outside lab take some time to get there? This can allow some growth even if the sample was cold. Did both labs use the same method ?

Bacteria are rarely distributed evenly in any sample. Unless the variation is very large, it is normal to see variation within the same sample. Microbiology isn't like chemistry !

I agree with Cathy.

Even something as basic as are they using the same agar and incubation temperatures are important, or how the sample has been handled during transportation.

No 2 samples ( in Micro or chemistry!) will give you an exact same reading. You also have to allow for human error (at either lab, either contamination or incorrect dilution factor or that the sample isn't homogenously mixed)

Caz

[poppysnoss]

Totally agree with the others.

What kind of foodstuff is being analysed?
Is it shelf stable or chilled?
What size of sub-sample are you taking for analysis and what are you sending of?
What types of tests are being carried out? Assumably indicator organisms?
Again, what methods?

An example is E. coli - many labs use pour/spread plate on TBX agar then incubate at 44 degees C for 24 hrs. Many other labs incubate at 37 degrees for 4 hours first to minimise stressed cells , then transfer to 44 degrees C.

However, another very common method is the membrane method, which in my experience is a much more sensitive test and I always get higher recovery when analysing the same inital suspension using both methods.

Anyway, just to give you a few examples...

Assumable you are analysing prior to the extrnal lab. If microorganisms are present beforehand and obviously depending on conditions, they may have scope to grow, therefore giving a higher result later.

Give us more information.

[Chac]

The product wich has been analyzed is a liquid sugar with a Brix about 68°.

One person from our laboratory staff took a 1000ml sample from one tank in one step. 500ml from this sample has been analyzed in our laboratory and the other 500ml was send to the external lab.. The sample was send in a carton with the common german post system. What happend on the way to the lab I can't tell but the sample couldn't get to warm because it's really cold outside.

The sample has been analyzed in both lab with the same standardized methods: IFU 2;3;4. These methods are commonly used in the sugar industry (well, at least in Germany).

Between our results and the results from the lab lay 12hours. They found 50times more yeast than we did.

I hope this will help a bit more.

@ Charles:

Nice you remember me . I've been on vacation the last 2 1/2 year *only kidding* No, the truth is, that I've worked in a chocolate factory but in the product developing. During this time I wasn't dealing with food safety neither did I have the opportunity nor the time to deal with this stuff, nearly working all day long. In the end it was too much to take so I quit and now at my new work I'm back into the "subject". I'm still working in the food developing but also in the quality section because both parts are quite linked with each other.

[Tony-C]

The product wich has been analyzed is a liquid sugar with a Brix about 68°.

The sample was send in a carton with the common german post system. What happend on the way to the lab I can't tell but the sample couldn't get to warm because it's really cold outside.

Between our results and the results from the lab lay 12hours. They found 50times more yeast than we did.

Any micro samples you send should be refrigerated or in a cool box that is capable of maintaining refrigeration temperatures.

We used to data log the temperature of a dummy sample sent in the same package if we encountered any problems.

Regards,
Tony

[Chac]

Any micro samples you send should be refrigerated or in a cool box that is capable of maintaining refrigeration temperatures.

I've talked with our quality manager about this and she explained me that this is not really necessary because of nthe high amount of sugar and with the temprature outside yeast don't increase so fast. If the temperature in the sample don't rise over 25°C it's allright. In fact if the sugarliquid gets to cold cristallization will starts.
It's always done in this way and there never has been such troubles. Of course the results differ from lab to lab but the problem is the high of the difference.

One thing that we can imagine is that the bottle in wich the sample was send was not totally clean. But our quality manager is not really satisfied with this suggestion.

[cazyncymru]

I've talked with our quality manager about this and she explained me that this is not really necessary because of nthe high amount of sugar and with the temprature outside yeast don't increase so fast. If the temperature in the sample don't rise over 25°C it's allright. In fact if the sugarliquid gets to cold cristallization will starts.
It's always done in this way and there never has been such troubles. Of course the results differ from lab to lab but the problem is the high of the difference.

One thing that we can imagine is that the bottle in wich the sample was send was not totally clean. But our quality manager is not really satisfied with this suggestion.

Actually yeasts can grow at 5 degrees.

[FSSM]

I've talked with our quality manager about this and she explained me that this is not really necessary because of nthe high amount of sugar and with the temprature outside yeast don't increase so fast. If the temperature in the sample don't rise over 25°C it's allright. In fact if the sugarliquid gets to cold cristallization will starts.

I totally agree, cristalization would reduce osmotic pressure, could your yeast be osmotolerant ones?

Have you checked the containers in wich you sample? Shouldn´t a contaminated container show other micro parameters deviated also?

Regards,

FSSM

[Charles.C]

Dear Chac,

Chocolate sounds a nice occupation also. I remember another poster did this and the company offered free samples every month to reduce pilferage.

50x seems high (the absolute numbers are relevant also, eg 1,50 or 50,2500?) but so does transport at 25degC also. Tonys comments are pretty typical IMEX though sugar not my field.

I usually add one dummy sample with known characteristics if the analysis is particularly important or unusual. May be expensive but sometimes worth it.

Other results (???) in line or not ?

Rgds / Charles.C

[sudarshan]

One thing that we can imagine is that the bottle in wich the sample was send was not totally clean. But our quality manager is not really satisfied with this suggestion.

Hi Chac

But i am not agree with Quality manager...because sampling is as mush important as test the sample...if sampling goes wrong....everything goes wrong

Edited by sudarshan, 16 February 2010 - 10:17 AM.

[Charles Chew]

A Food Safety System does NOT guarantee safe food. Laboratory testing does NOT guarantee accuracy of results .... hence "Uncertainty in Measurement"

[Tony-C]

I've talked with our quality manager about this and she explained me that this is not really necessary because of nthe high amount of sugar and with the temprature outside yeast don't increase so fast. If the temperature in the sample don't rise over 25°C it's allright. In fact if the sugarliquid gets to cold cristallization will starts.
It's always done in this way and there never has been such troubles. Of course the results differ from lab to lab but the problem is the high of the difference.

I don't agree with this but you can validate this claim by arranging to test duplicate samples after holding
1. At 5 C for 12 hours
2. At 25 C for 12 hours

Regards,

Tony

[Sujit]

I think the exercise should be repeated and the collected sample well shaken for uniformity and then immediately distributed into two 500 ml bottles.
In my experience with external accredited lab, our results were with in 10%, which was acceptable.
Regards,
Sujit

[Jon5]

I have a few notes to add.

• At this high a degree brix, temperature does not matter. Yeast will not actively grow under these conditions. If they do survive, they've gone dormant - many yeasts will sporulate.
• Is there a dilution of the product and an agitation step as part of the analysis? Cells will typically "group" together, and must be broken apart to achieve an effective cfu count.
• Do you use any sort of sanitizer in your incubators? Quaternary ammonia compounds and other chemicals commonly used can actually volatilize and get into the media, and inhibit growth. Also, if you're washing, sanitizing, and re-using any other equipment (glass petri dishes, etc) this can occur.
• Do you use a hot water bath? The condition/contents of the water may have a result.
• Verify that incubation conditions are exactly the same - including ambient gas in the incubator.
• If you're making your own media, there are a whole lot of other variables that can enter the equation as well.

Hope this helps. Yeast are close to my heart. :-)

[AS NUR]

IMO the reason why your result so differents is reproducibility of Micro analysis is high enough, so you have to consider "uncertainty value" as Charles Chew said. and the unceratinty came from : Analysts, LAb, equipment, methods etc.