Open Access
Issue
Vet. Res.
Volume 39, Number 1, January-February 2008
Number of page(s) 12
DOI http://dx.doi.org/10.1051/vetres:2007041
Published online 26 October 2007
How to cite this article Vet. Res. (2008) 39:03
Vet. Res. (2008) 39:03
DOI: 10.1051/vetres:2007041

Farm practices to control E. coli O157 in young cattle - A randomised controlled trial

Johanne Ellis-Iversen1, Richard P. Smith1, Steven Van Winden2, Giles A. Paiba3, Eamon Watson4, Lucy C. Snow1 and Alasdair J.C. Cook1

1  Centre for Epidemiology and Risk Analyses, Veterinary Laboratories Agency, New Haw, Surrey KT15 3NB, United Kingdom
2  Royal Veterinary College, Hawkshead Lane, North Mymms, Hertforshire AL9 7TA, United Kingdom
3  Food and Farming Group, DEFRA, 1A Page Street, London SW1P 4PQ, United Kingdom
4  VLA-Winchester, Itchen Abbas, Winchester, Hampshire, SO21 1BX, United Kingdom

(Received 5 February 2007; accepted 25 July 2007; published online 26 October 2007)

Abstract - A randomised controlled trial was used to investigate the effect of three complex management intervention packages to reduce the burden of E. coli O157 in groups of young-stock on cattle farms in England and Wales. All intervention farms were assigned measures to avoid buying in new animals and having direct contact or sharing water sources with other cattle. Furthermore, package A (7 farms) aimed to keep a clean environment and closed groups of young-stock; package B (14 farms) aimed for improved water and feed hygiene, whilst package C was assigned both A and B. The control farms (26 farms) were asked not to alter their practices. Farms, which were assigned intervention package A, exhibited a 48% reduction in E. coli O157 burden over the 4.5 months (average) of observation, compared to 18% on the control farms. The effect of package A compared to the control farms in a crude intention-to-treat model was RR = 0.26 (p=0.122). When the risk ratio was adjusted for actual application of the different measures, the effect of intervention package A became stronger and statistically significant (RR = 0.14 p=0.032). Statistical evidence (p< 0.05) showed that dry bedding and maintaining animals in the same groups were the most important measures within the package and weak evidence (p< 0.1) showed that a closed herd policy and no contact with other cattle may also be of importance. Compliance with the other measures in package A had no influence on the effect of the package. No evidence of effect of the other two intervention packages was found.


Key words: VTEC (STEC) E. coli O157 / randomised controlled trial / prevention / zoonotic control in cattle

Corresponding author: j.ellis-iversen@vla.defra.gsi.gov.uk

© INRA, EDP Sciences 2007