The STEC bacterium has been tested in milk sold through Swiss milk taps. Milk samples from more than 120 dairy farms were tested for STEC, among other things. STEC refers to the Shiga toxin-producing strains of E. coli. There are many of them. Young children (<5 years) are particularly susceptible, and it can lead to kidney failure, kidney transplantation and even death. And nobody wants that. 

Milk taps sampled

The combination of a positive PCR test followed by the successful culturing of the bacteria is regarded as ‘proof’ of a positive STEC result. In a few milk samples (3.2%), genes of the STEC bacterium were found using a PCR test alone. The question then arises as to whether this constitutes a positive STEC result, hence the question mark in the title of this article.

It is striking that in a large proportion of the milk samples, raw milk is found with extremely high bacterial counts (Table 1; above). In 36 of the 124 samples, the bacterial count exceeds 100,000 cfu/ml. It was also found that a proportion of the samples contained high numbers of Pseudomonas, cold-loving bacteria that proliferate in cold milk during storage. The authors write: “In total, 15 samples showed a bacterial count above 1 × 10⁶ CFU/ml, including four samples with a count of more than 1 × 10⁷ CFU/ml. In these samples, the bacterial flora was predominantly dominated by psychrotrophic Pseudomonas species.”

This is likely an indication that dairy farmers are leaving ‘their fresh milk’ in their taps for too long. When looking at results from, for example, German Vorzugsmilch, bacterial counts above 10,000 cfu/ml are virtually non-existent. Similarly, the milk supplied to the Raw Milk Company (De Lutte, NL) for the production of raw milk kefir tends to fall between 3,000 and 10,000 cfu/ml.

Table 1. Classification of samples into 3 classes for bacterial count (<10,000; 10,000–100,000; >100,000 cfu/ml), (top) or E. coli concentrations (0; 10–100; >100), (bottom)

		Classes for total germs (cfu/ml)
Values (x 1,000 per ml)	<10	10-100	>100
Number of samples	36	52	36
Bacterial count (CFU/ml *1,000 (via log 10))	3	21	803
E. coli/ml (log 10)	0.6	3.8	2.9
Mean of STEC	2.8%	3.8%	5.6%
		Classes for E. coli (cfu/ml)
Values (per ml)	<10	10-100	>100
Number of samples	90	23	11
E. coli/ml (log 10)	0	25	764
Bacterial count (CFU/ml *1,000 (via log 10))	32	24	149
Mean of STEC	2.2%	4.3%	9.1%

There are also milk samples with a significantly elevated concentration of E. coli (Table 1; below). This could primarily be attributed to poor pre-treatment and cleaning of the teats; too much manure (containing E. coli) is entering the milk. In 11 of the 124 samples, the E. coli value is >100 cfu/ml. By dividing the farms into three classes (low, medium, high), one can estimate what the likelihood of a higher bacterial count or higher E. coli levels means for the risk of STEC. STEC increases with both a higher bacterial count (from 2.8% to 3.8% and then to 5.6%), but particularly with a higher number of E. coli in the milk (from 2.2% and 4.3% to 9.1%). It is striking that in samples where no or virtually no E. coli is found (result E. coli <10), there is still a 2.2% chance of a positive PCR result for STEC.

In their conclusion, the authors write: “The detection of Campylobacter jejuni, STEC and L. monocytogenes in samples of raw milk confirms the risk of exposure to foodborne pathogens associated with the consumption of untreated raw milk from these vending machines.” The article concludes with yet another warning to be cautious about raw milk consumption, as STEC genes have been found and the adage ‘where there’s smoke, there’s fire’ applies. However, they overlook their own definitions regarding whether a STEC should or may be termed a positive STEC result. A positive PCR result alone is apparently sufficient, but this contradicts international definitions.

We often see that a certain amount of PCR noise remains when looking solely at positive PCR results. In Switzerland too, there are 2.2% positive PCR results without E. coli or live STEC bacteria always being found. Previously, in the study of Gouda farmhouse cheese in the Netherlands, there was also a ‘noise threshold’ of 2% false-positive PCR results.

A valid conclusion that could have been drawn from this study is that Swiss dairy farmers need to clean their milk taps more thoroughly and frequently, and ensure that their raw milk is replaced in the taps daily or every two days. Milk cooling must also be properly managed. This would likely further reduce the risks associated with raw milk consumption, something we have long known from the way raw milk is offered and traded elsewhere in the world.

Conclusion

The finding of around 2% positive PCR results in milk samples without any live STEC bacteria being detected will continue to be a source of confusion in the STEC debate. Is such milk a hazard or not? This was also previously the case with Dutch Gouda farmhouse cheese. The question is whether warnings based solely on a PCR result regarding the danger of STEC in raw milk and raw milk products are always justified.

Literature

• Paravicini, T., Stevens, M. J., Barmettler, K., Cernela, N., & Stephan, R. (2026). Assessment of the Microbiological Quality of Raw Milk Sold Through Vending Machines at the Farm Level in Switzerland. Pathogens, 15(3), 322.