Log in or Register for enhanced features | Forgotten Password?
White Papers | Suppliers | Events | Report Store | Companies | Dining Club | Drinks | Retail | Videos
Food Technology
IT & software
Return to: FBR Home | Food Technology | IT & software

Why Count Bacteria on Surfaces?

Published 09 August 2016

Hygiena International continues with its series of weekly articles that discuss monitoring microbiological contamination on surfaces. This week features its second publication, which carries on the topic point by querying, "Why Count Bacteria on Surfaces?"

In the previous article we discussed the limitations of microbial measurement when significant numbers were present in the sample. Even the best laboratories running several replicates in proficiency test schemes are expected to see significant variation between laboratory results on the same sample which can be 1 log (10 fold) but in practice the results for indicator organisms such as coliforms and enterobacteriaceae are worse (3-4 log variation has been observed). So what can we expect from environmental samples where the microbial load is much lower?

The challenges are many fold, so that enumerating microbial contamination in environmental samples is of limited value and we need to consider a different approach.

Environmental monitoring is required to measure the presence of total bioburden (Total aerobic bacteria), indicator organisms of poor hygiene such as coliform and enterobacteriaceae, spoilage organisms such as lactic acid bacteria or yeast, and pathogen bacteria such as listeria and salmonella. There are no standards for environmental samples due to the varied and unique nature of each manufacturing facility and process.

For many prepared foods it is generally accepted that pathogens should be absent, indicator and spoilage organism should be very low ( <10 or <100 cfu) or below a detectable limit and the total bacteria count has broad guidelines of <10cfu as good and >1000 cfu as unacceptable. The closer the environmental sample test location is to the open product, then the tighter the specification since there is a greater risk of cross contamination.

Environmental monitoring is typically conducted after cleaning to verify sanitation procedures. Similarly, in-process samples are tested to monitor the cross contamination hazard or build-up of contamination during manufacture.

Accordingly, it is expected that microbial contamination will be low. Difficult to clean surfaces are usually used as test locations but microbes are not evenly distributed even on flat surfaces. Microbial distribution on surfaces is often described as contiguous (see image below) so obtaining a representative sample is very difficult and the probability of detection is low.

As discussed previously, microbiological enumeration methods lack precision and the uncertainty of measurement is compromised further by the low numbers. Plate counts should have a minimum of 25 colonies to give a reasonable probability of enumeration (25% error). As the number of colonies per plate decreases so the standard error of the result increases to 50% for 5 colonies and 100% for 1 colony, so the enumeration in the range of 1-10 colonies per plate is very unreliable.

Validation studies for surface contamination usually use factorial designs where 20 replicate samples are used and the percentage positive samples are recorded and not the number of colonies. Since most environmental samples typically yield results of 1-10 colonies or fewer then enumeration has a high inherent variability and begs the question as to the value of the information generated. It would be better to test a greater number of samples and express the results differently in order to get a better understanding of the levels and distribution of contamination and to trend these over time rather than rely on a single random one-off determination and record the number of colonies presentas an absolute number.

Using a statistical ‘binning’ technique is a better way to analyse data from surface contamination and to compare methods. This smooths the variation due to sample distribution and within the method itself to improve the correlation from <60 to >90% and give greater confidence in the result.

You can get a better understanding of surface contamination and risk if you change the perspective and the yardstick.



Source: Company Press Release