Shelf life refers to the length of time a food remains palatable - from the time it is harvested or processed until it is eaten. The designated shelf life of a product is an integral part of a food safety program. The aim in performing a shelf life evaluation of a food product is to ensure the food quality throughout the stated use-by or best-by period. Food quality falls into two sections; food spoilage and safe food. Food is considered spoiled when an undesirable change in the colour, flavour, odour or texture has occurred. A product can be perfectly safe to eat but due to spoilage has lost its appeal to a potential customer. A product can also contain physical, chemical or microbial elements that can cause harm to the consumer and therefore represents an unsafe food. This discussion will deal with the microbial elements of shelf life evaluations.

When designing a shelf life evaluation seven questions need to be answered; how long should the evaluation period be, how should the samples be held during the evaluation period, what microorganisms need to be tested, how often during the evaluation period should samples be analysed, how many samples should be analysed at each testing interval, what type of sample should be collected and how are the results interpreted.

The following discussion continues from BioTalk 18.

How many samples should be analysed at each testing interval?
There are three factors to consider when deciding on the number of samples to be analysed at each individual testing interval.

Firstly, microbes are not evenly distributed through a food matrix. Irrespective of the degree of mixing, when this is applicable, the number of  microorganisms in each gram differs throughout the product and even more so throughout a batch of products. Unless it has been proven by laboratory testing that homogeneity of microbial counts does occur through out a batch of products, it can be erroneous to presume that the result of one sample unit from a batch is representative of the whole batch of products.

Secondly, microbiology is not an exact science. The whole process of the culture of bacteria and fungi requires sampling and dilution steps that decrease the accuracy of the final result reported. There are acceptable errors in the accuracy of balances used to weigh the initial sub-sample (viz. 10g could be 10.1 or 9.9 grams), or the weight of dilution fluid used to make up the initial one-in-ten dilution. An acceptable accuracy of a pipette used to perform serial dilutions of the initial one-in-ten dilution or agar plate inoculation can be 2%. There can be up to 6 or more pipetting steps alone for some quantitative counts. When all these variations are added together we can easily find a 10 to 15% variation of duplicated samples of identical microbial levels.

Add to this the fact that microorganisms are not evenly distributed within a food matrix as already discussed and it is frequently found that a result of two results from the same food sample can quite reasonably be within a log phase separation. For example, a plate count of 240,000 CFU/gram and 710,000 CFU/gram would not necessary be regarded as dissimilar.

The third factor to be considered when deciding on the number of samples to be analysed at each individual testing interval is cost. Whereas taking 10 samples at each testing interval and using the median or average count would provide the most reliable data this is rarely economical. The decision needs to be based on a combination of the variables discussed and the total cost of the shelf life trial. In some instances a pilot study of one analysis, such as plate count, may provide the necessary information to select a suitable number of samples to be analysed at each testing interval.

What type of sample should be collected and how are the results interpreted?
The type of sample will depend on the likely source for microbial growth and contamination. For most food products including mince meat, liquid products, cereal and processed foods a core sample is required.

For products such as whole fruit and vegetables, a surface sample or surface rinse sample may be required. The contamination of these types of products can often occur during the post harvesting washing procedure using contaminated water. However in some cases organic fertilisers or contaminated irrigation water can be a source of faecal microorganisms which may be absorbed into the food product in addition to the surface. In these situations both forms of sampling may be indicated.

Interpretation of the shelf life results are based on the Food Standards Code, local government and/or international microbial food guidelines. These values need to be determined before the initiation of a shelf life trial.