Analysis of the baseline survey on the prevalence of Salmonella in holdings with breeding pigs, in the EU, 2008 - Part B: factors associated with Salmonella pen positivity


European Food Safety Authority
EFSA Journal
EFSA Journal 2011;9(7):2329 [159 pp.].

EFSA wishes to thank the members of the Task Force on Zoonoses Data Collection: Marta Bedriova, Susan Chircop, Georgi Chobanov, Veronica Cibin, Jürg Danuser, Kris De Smet, Kostantins Golubevs, Matthias Hartung, Birgitte Helwigh, Merete Hofshagen, Simona Iannetti, Sarolta Idei, Patrícia Inácio, Elina Lahti, Lesley Larkin, Emma Martín Denia, Peter Much, Edith Nagy, Ioana Neghirla, Lisa O’Connor, Jacek Osek, Manca Pavšič, Christodoulos Pipis, Saara Raulo, Jose Luis Saez Llorente, Julien Santolini, Petr Šatrán, Snieguole Sceponaviciene, Joseph Schon, Richard Smith, Jelena Sõgel, Johanna Takkinen, Chris Teale, Kilian Unger, Rob Van Oosterom, Luc Vanholme, Dimitris Vourvidis for the endorsement of the report; the members of the working group: Vojislava Bole-Hribovšek, Mariann Chriél, Rob Davies, June Fanning, Arjen W. van de Giessen, Laura Pérez Palancar, Antonia Ricci, Nicolas Rose, Lucy Snow for the preparation of this EFSA scientific output; Arjan Stegeman for the peer review; the EFSA’s contractors: Hendriek Boshuizen and Gerda Doornbos from National Institute for Public Health and Environment (RIVM), Tine Hald from Danish Technical University, National Food Institute, Elinor Jones, Mark Arnold and Alasdair Cook from the Veterinary Laboratories Agency (VLA), and the EFSA’s staff members: Giusi Amore, Francesca Riolo, José Cortinas Abrahantes and Frank Boelaert for the support provided to this EFSA scientific output.

Scientific Report of EFSA
On request from
European Commission
Question Number
25 maggio 2011
26 luglio 2011
European Food Safety Authority (EFSA) Parma Italy
Download Article (4.04 MB)

A European Union-wide Salmonella baseline survey was conducted in 2008 in holdings with breeding pigs. A total of 1,609 randomly selected holdings housing and selling mainly breeding pigs (breeding holdings) and 3,508 holdings housing commercial breeding pigs and mainly selling pigs for fattening or slaughter (production holdings) were sampled. In each selected holding, pooled fresh faecal samples were collected from 10 randomly chosen pens of breeding pigs over six months of age, representing the different stages of the breeding herd, and examined for the presence of Salmonella. Analyses at country-level demonstrated a strong positive association between the prevalence of Salmonella-positive breeding holdings and the prevalence of Salmonella-positive production holdings, suggesting a vertical dissemination of Salmonella between the holdings. Based on the combined results from breeding and production holdings, multivariable regression analysis showed that the odds of Salmonella-positive pens with pigs increased with the number of breeding pigs in the holding and with the following pen-level factors: flooring systems other than slatted floors or solid floors with straw, presence of maiden gilts, number of pigs per pen, feed of commercial compound origin or pelleted feed. A tendency towards some Member State group-specific Salmonella serovars was identified, but spatial distribution of other serovars was heterogeneous. S. Typhimurium and S. Derby were widespread and dominant in the EU, in both breeding and production holdings. However, many other serovars were relatively prevalent in Western EU Member States. A complementary within-holding prevalence study indicated that, due to a non-perfect diagnostic test sensitivity, the observed EU-level prevalence of Salmonella-positive holdings with breeding pigs was roughly 80% of the estimated true EU-level prevalence. But this proportion varied between Member States.


A European Union-wide baseline survey on Salmonella in holdings with breeding pigs was carried out in 2008. In the context of this survey, breeding pigs were defined as sows and boars of at least six months of age kept for breeding purposes. The survey distinguished between breeding holdings (holdings housing breeding pigs and delivering replacement breeding pigs to breeding holdings and production holdings) and production holdings with breeding pigs (holdings housing breeding pigs and producing mainly pigs for fattening or slaughter). Breeding and production holdings were randomly sampled from holdings harbouring at least 80% of the breeding pig population in each Member State. In each selected holding, pooled faecal samples for Salmonella detection were collected from 10 randomly selected pens of breeding pigs representing the different stages of production of the breeding herd (maiden gilts, pregnant pigs, farrowing and lactating pigs, pigs in the service area, or mixed). A total of 5,117 holdings with breeding pigs with validated results from 24 European Union Member States, plus Norway and Switzerland, were included in the survey analyses, corresponding to information on 1,609 breeding holdings and 3,508 production holdings. Samples were taken from a total of 48,951 pens selected in both breeding and production holdings. The results of the analysis of Salmonella prevalence have already been published by the European Food Safety Authority on 17 December 2009 in the Part A report. The present Part B report provides the results from analyses of the associations of 19 pen- or holding- level factors and Salmonella positivity of pens in holdings with breeding pigs. The investigated prevalence was the observed prevalence, meaning that the prevalence estimates did not account for imperfect test characteristics. Also the results from correlation analyses between Salmonella prevalence in breeding and in production holdings, from analyses of the Salmonella serovar distribution across the European Union, and from analyses of an additional within-holding prevalence study carried out by five Member States are also presented in this part B report.

Correlation analyses at country-level demonstrated a strong and significant positive association between the prevalence of Salmonella-positive breeding holdings and Salmonella-positive production holdings, suggesting the likelihood of vertical dissemination of Salmonella between the holdings. This hypothesis was further underpinned by the significant positive association between the prevalence of Salmonella-positive holdings with breeding pigs of the present survey and the 2006 to 2007 baseline survey prevalence of Salmonella-positive slaughter pigs.

Multivariable regression analysis of the combined dataset for breeding and production holdings showed that at European Union[1] level the odds of pens being positive to Salmonella increased as the holding size increased. Also the holding gilt replacement policy (way in which gilts are replaced in the holding) was associated with Salmonella pen positivity in breeding holdings, but not in production holdings. The odds of pens being positive to Salmonella increased with the number of pigs in the pen, with a 3% increase in odds per 10 additional pigs. The production stage of the pigs was also found to be significantly associated with Salmonella positivity with pens containing maiden gilts having higher odds than pens with pregnant or farrowing and lactating pigs. In breeding holdings, pens with fully slatted floors were associated with lower Salmonella-positivity than the category of pens with ‘other’ floor type. In production holdings, fully slatted pen floors had lower odds of being positive than outdoors in fields or paddocks as well as the other types of floor, except ‘solid floor with straw’ and pens where the floor type was classified as ‘other’. Pens where pigs were fed with feed of commercial origin had higher odds of Salmonella positivity compared to those in which either home-mill mixed feed or feed from some other sources were used. Also, pens where pigs received pelleted feed as type of diet were associated with a higher Salmonella-positivity when compared to pens where pigs were fed with meal or wet feed.

In addition, the odds of pens being positive with Salmonella varied significantly between countries and between holdings within a country, even when other associated factors were accounted for. Moreover, swab samples (swab passed through accumulated mixed faeces in the pen) were more likely to be Salmonella-positive than composite samples (comprising individual pinches from faeces), suggesting that sampling using swabs was a more sensitive method for detecting Salmonella in the pen than the composite sample.

Holding level factors that were included in the analysis but which were not significantly associated with Salmonella-positive pens were season of sampling, delay between sampling and testing, type of breeding or production holding, and boar replacement policy. Non-significant pen-level factors were age category of the pigs, gender of the pigs, indoor/outdoor production, individual housing, all–in/all-out production and cleaned, feed/water supplement and the use of antibiotics in the pig pens. However, for some of the factors the power of the analyses was low due to too few samples in some specific categories. Moreover, the analyses showed that 56% of the unexplained variance in the Salmonella-positive pen results might have been attributable to holding-specific factors for which no data were gathered during the survey and/or to the clustering of Salmonella linked to its infectious character.

The highest estimated theoretical reduction of Salmonella-positive pens would be observed if specific control measures were put in place that focus on the reduction of the exposure to feed of commercial compound origin and pelleted feed diet.

A notable variation in the number of different Salmonella serovars was observed across the European Union Member States indicating a heterogeneous serovar distribution between participating countries. S. Typhimurium and S. Derby were widespread and dominant in most Member States, while other serovars, such as S. London, S. Infantis or S. Rissen were frequently isolated in some specific countries and their relevance cannot be generalised to the European Union as a whole. Many serovars isolated in the breeding pigs’ survey are also common in slaughter pigs as well as in other food producing animal species and food thereof, indicating that the potential for contribution of these serovars to human infections may be shared between different sources.

The analysis of a complementary within-holding prevalence study allowed estimating the sensitivity of the pooled faecal sampling method, as well as the EU and MS level true prevalence of Salmonella-positive holdings with breeding pigs. The sensitivity of the pooled faecal sample was estimated to be 92% and it was shown to increase with the prevalence of positive pigs within the pen. Moreover, the results indicated that the EU level true prevalence of Salmonella-positive holdings with breeding pigs, as reported in the Part A report, and based on the sampling of 10 pens per holding, could be underestimated by 20%, although this percentage would vary between the Member States. It is recommended that Member States consider the factors found to be associated with Salmonella-positive pens at the European level in this survey, when they are designing and implementing national Salmonella control programmes for breeding pigs. Further national studies identifying more closely the factors that put pens with breeding pigs at risk of becoming infected with Salmonella in a country are recommended, taking into account the national Salmonella prevalence and the characteristics of the national breeding pig population. Also national investigations on prevention and intervention measures to contain Salmonella and achieve Salmonella reduction in holdings with breeding pigs are recommended. Since risk factors may vary between Member States and/or serovars, Member States are also encouraged to conduct serovar-specific analysis using their country specific data in order to identify risk factors for relevant serovars within their own country. Member States are encouraged to develop and enhance Salmonella controls in breeding holdings because these holdings have a unique potential role in the dissemination of Salmonella contamination throughout the whole production chain, as well as in contamination of the environment. Pooled faecal samples proved to be a robust and economic sampling method for surveys and should be used in future studies, as well as for monitoring the Salmonella status of breeding herds. Sampling procedures require standardisation to enhance sensitivity and comparability of monitoring results. Those Member States that did not participate in the within-holding prevalence study may wish to conduct their own research to validate pooling in their own situations.

Salmonella, pigs, breeding pigs, baseline survey, risk factors, EU
Print on demand
Number of Pages