25 October, 2007 (United Kingdom)—European Food Safety Authority (EFSA) is striving hard to obtain uniformly low salmonella levels in layer flock holdings across all European Union (EU) member states. Final version of their 2005 study released this year showed that there is a long way to go to achieve an acceptably low level for the Community as a whole and to bring under control huge differences between countries. Deficiencies in design of the study including sampling methods may be hiding higher overall salmonella levels and distorting differences between member states.

Tests were made on dust and other materials in poultry houses, as well as bird faeces, to assess levels of Salmonella spp on holdings. Sampling and testing was carried out on 1486 holdings across the EU to give a Community-weighted observed holding prevalence for salmonella of 30.8%. Huge differences between countries existed within this average figure – from a minimum 0% in Luxembourg and Sweden to a maximum of 79.5%, 76.2% and 73.2%, respectively, in Portugal, Poland and Spain.

Salmonella enteritidis and S. typhimurium are usually the two most common serovars in the region. S. enteritidis is responsible for more than half of all human food poisoning cases within the EU. The two serovars were present on 986 holdings to give an observed holding prevalence of 20.4%. Again, the average figure hid wide country differences, ranging from 0% in Ireland and Sweden to 62.5%, 55.5% and 51.5%, respectively, for Czech Republic, Poland and Spain.


Salmonella was found more frequently on holdings with large layer flocks. It appears this was not taken into account in the method of sampling used and may have affected the results. Number of samples taken (5 faecal samples and 2 dust samples) was constant, irrespective of flock size. Flocks ranged from 1000 (minimum size to qualify for testing) to 100,000 birds.   

Another feature of the survey questions was the testing of just one flock per holding, irrespective of how many flocks that holding contained.  This suggests the true holding salmonella prevalence may have been higher than the observed figure. Multiple-flock holdings that showed negative from the single flock tested could have one or more additional flocks that were positive.

EFSA’s report accepts this anomaly could have contributed to an under-estimation of the salmonella prevalence but emphasised that the figure obtained was generally higher than recorded in the national zoonoses reports submitted by member states. However, individual member states use different sampling methods, ruling out any worthwhile comparison with each other or the EU average calculated in such a report.