Different bacterial count in two repeats of the same sample

Good day, everyone!

My name is Agne, I am new in this forum, but really need your help.

In our company we do some microbiological tests and when doing enterobacteriaceae counting we get two completely different results even though we test the same sample. One day we get less than 10 CFU, other day - more than 600 CFU. 

Maybe you have any ideas what could cause this?  :unsure:

Are you taking 2 swabs from the exact same place one after the other?

Good day, everyone!

 

My name is Agne, I am new in this forum, but really need your help.

 

In our company we do some microbiological tests and when doing enterobacteriaceae counting we get two completely different results even though we test the same sample. One day we get less than 10 CFU, other day - more than 600 CFU. 

Maybe you have any ideas what could cause this?  :unsure:

Hi Agne,

Unfortunately there are soooo many possible causes. Microbiological counts are also notoriously inaccurate.

You will probably need to clarify various things like -

type of product ?

method sampling ?

procedure used ?

same operator/equipment ?

As per previous post, it's unclear what you mean by "same sample". Is it a homogeneous  liquid ?

I asume you are doing other counting measurements, eg aerobic plate count. What kind of reproducibility do their results give you ?

Thank you for your reaction.

No. We did the test one day and when counted the CFU and got a very large number, then redone the test again and then got a lot smaller number. But there should be such a big difference between two results, I quess..

AgneP, was there any differences in environment or equipment between the first and second test? Possibly if a cleaning had been conducted between or the samples were in 2 different containers for example?

First of all sorry for my English, I am not the best English writer. 

The type of product is milk powder. 

We do the liquid solution and then pour it in the petri dish and cover with medium. 

The same operator did it with the same equipment. 

It is strange because we never had such similar problems.. and it is scary, because what if when we get small result it is in reality should be bigger?? 

Yes, we also do total bacterial count and moulds and yeasts. 

The sample was in the same bag, no changes in environment have happened between these two tests.. 

Hi Agne,

Yr English is fine. :smile:

Was the product in bag all from same production lot ?

i assume this is an ISO-type procedure using VRBG agar + 1ml of a decimal dilution of milk powder (eg 25g/225ml diluent), ie no colonies on plate means <10cfu/gram.

So 60 colonies > 600cfu/gram

(It may depend on yr spec. but strictly such low numbers are not recommended for routine plate counting. An  MPN procedure would be more precise but more work/slower).

Do you use  duplicate plates ? If so what were the pairs of results for each analysis ? (this acts as a partial check for the operator/sample/procedure/equipment)

Have you tried repeating yr 2-day exercise on a different bag ?

I would suggest that perhaps take a much larger sample, ensure it is mixed very very well then take the sample size you need as a representative of the product.

I would suggest repeating the 2 tests again to see if you get the same results...............as Charles mentioned, there are so many factors involved that it would influence the result its nearly impossible to narrow it down (unclean hands, ecoli CFU higher in a pocket of powder, hose touch the floor, etc)

Thanks everybody!

We are redoing the test one more.

About lot, yes the bag is from the same lot :) 

And about the procedure, you are absolutely right, we use this ISO type method. 

The idea abou two plates from the same sample is very good! 

Thanks! 

Hi AgneP,

As people have explained above, there are a number of possible causes.

• The powder is contaminated, but not homogeneously. 
• The sample was contaminated in the lab
• You have a rogue result

Is it your own lab or external? How well do they work? Are they ISO17025? Was the second result their retest or from a resubmitted sample? My experience of working in a micro lab is that errors can occur. However, the usual approach is that the powder would be reconstituted using 10g powder and 90g of diluent (could be BPW or Sodium Citrate depending on your requirements). About 100g of the rest of the powder is then put aside for up to 1 week to allow retesting if unusual results are found. From the 100g reconstituted material, 1g is plated and incubated. Unfortunately, this does not always give representative results as the powder is unlikely to be fully mixed.

Plating from the retained sample will not necessarily give the same result as the sample is taken from a different set of powder, hence the different results.

Plating 2 samples will limit the effects of contamination in the lab and/or rogue results. 

However, you can never be sure that there was no contamination as your original sample has been use up.

Isn't life fun!

Micro plating can vary widely even with the same person and the same sample.  If you see results within 1.5 log of one another on the same sample with same person or different persons then you can rest assured your micro plating is consistent.