Microbial Quality of Unrefrigerated Desserts

January - March 1997

Report prepared by Simon Rockliff and Geoffrey Millard

Samples collected and analysed by the staff at ACTGAL


1.1 The objective of the survey was to determine the bacteriological status of Unrefrigerated Desserts in the ACT.

2. Background

2.1 Previous surveys have found that food pathogens have been isolated from this product type when stored at room temperature. This survey was undertaken to determine if this product type continues to represent a public health risk. The Australia and New Zealand Food Authority (ANZFA) Food Standards Code (FSC) have no standards for this product type.

3. Survey

3.1 The survey was conducted between January and March 1997 with a total of 68 samples of various unrefrigerated desserts being collected from 32 different outlets. Samples were purchased as consumer items over the counter by the ACT Government Analytical Laboratory (ACTGAL) staff and analysed by the Microbiology Unit. The samples were assessed for hygiene quality by the Standard Plate Count (SPC), E. coli and coagulase-positive Staphylococcus ( Coag+staph ) analyses and for specific food pathogens such as B. cereus and Salmonella spp. The survey consisted of the collection of multiple samples once from most outlets in the ACT. A vanilla slice sample was repeated from one outlet due to unacceptable results being found.

3.2 Table 1 gives the surveys acceptability criteria.

Table 1


Test Criteria








E. coli












Salmonella spp

Not detected*



# Units expressed in terms of Colony forming units (cfu) per gram. * Organism not detected in 25 g.

3.3 The maximum temperature occurring in Canberra on the day the samples were collected was sourced from the Canberra Meteorological Office in an attempt to correlate day time temperatures with microbiological results.

4. Results

4.1 The samples consisted of 24 Vanilla Slices, nine Custard Tarts, six Eclairs, two Long-johns, three Profiteroles, two Amorettos, two Custard Puffs and 20 single items. Results are expressed as colony forming units per gram (cfu/g) except for Salmonella spp which is expressed as isolated/ not isolated per 25 grams of analysed material.

4.2 The overall isolation rate for E.coli is given by Chart 1. E.coli was isolated from seven (10.3%) of the samples at a range of 21,600 cfu/g. Only one sample, a vanilla slice, exceeded the acceptability criteria for E.coli by being greater than 1000 cfu/g.

4.3 The overall isolation rate for Standard Plate Count is given by Chart 2. An acceptable SPC was determined as being less than 1,000,000 cfu/g and then either in the good or poor category depending on its SPC level. 37% of the samples had acceptable but poor SPC meanwhile, 20 (29%) of the samples had an unacceptable SPC. Analysis of the raw data did not reveal any noticeable trend of unacceptable SPC being related to any particular dessert type.

4.4 Coagulase-positive Staphylococcus was only isolated from two samples at levels of 50 and 300 cfu/g. All samples met the acceptability criteria of less than 1000 cfu/g.

4.5 Salmonella was not isolated from any of the samples and met the acceptability criteria of absence per 25 gram of analysed material.

cfu/g = colony forming unit / gram material analysed



4.6 The isolation rate for B.cereus is given by Chart 3. B.cereus was isolated from 15 (22%) of the samples. It was isolated from 11 of the 25 (44%) vanilla slice samples with a range of 200540,000 cfu/g. Eight (11.8%) samples exceeded the acceptability criteria for B.cereus of 1000 cfu/g. The vanilla slice results are skewed as a number of samples were repeats collected from one outlet because of unsatisfactory results. In samples other than vanilla slices, B.cereus was isolated from 2 of 9 (22.2%) custard tarts (700 and 1500 cfu/g), one eclair at the count of 1,400 cfu/g and one jam tart with a count of 400 cfu/g.



4.7 Temperature on day of collection is shown in Graph 1.

Graph 1



5. Discussion:

5.1 Analysis of the raw data did not reveal a correlation between the hygiene quality tests of SPC, E.coli or coagulase-positive Staphylococcus and the pathogens B.cereus or Salmonella spp. Samples with high Standard Plate Counts often didnt contain any pathogens and the reverse was also true.

5.2 The E. coli, coagulase-positive Staphylococcus and Salmonella spp levels for most of the samples was acceptable. Of some concern is the presence of B. cereus in 44% of the vanilla slices and 29% of samples having unacceptable levels of SPC. B.cereus can be a naturally occurring contaminant of the ingredients used to prepare some types of dessert and B.cereus and SPC organisms can proliferate if the product is not handled, stored or transported correctly. Evidence that some premises are not preparing, storing and/or transporting their desserts correctly is supported by the unacceptable B.cereus and SPC results obtained for different dessert samples, collected at the same time from these premises. Unacceptable B.cereus and E.coli counts in one sample warranted repeat sampling from that premise and subsequent follow up by the Health Protection Service Environmental Health Officers.

5.2 There was a suggestion that the high B. cereus counts on the 3rd of March, where caused by heat wave occurring at that time. Graph 1 demonstrates the ambient temperature on that day was slightly lower than average. If temperature was not the cause then inadequate preparation, transport, handling and/or storage was most probably responsible for the elevated B.cereus result in the vanilla slices.

6. Conclusion

6.1 There is no obvious correlation between bacterial indicators of hygiene quality and pathogens. The SPC, E. coli and B. cereus results indicate a problem associated with inadequate practices for storing and handling sampled desserts. This problem appears to be premises specific as detection of pathogens and/or quality indicator organisms was predominantly restricted to multiple samples collected from the same premises.