Speaking at the International Production and Processing Expo in Atlanta, Georgia, Dr Hurd also said that the risk of antibiotic resistance being passed down from the farm to humans was negligible and was no reason to stop the use of antimicrobials on the farm.

He said that the major problem concerning multi-drug resistant (MDR) salmonella was in the merging of "three scary features" by the media and campaigning organisations - the words resistance and salmonella and the political agenda of the organisations.

The number of incidents of has been shown to be about the same as the incidents of non-resistant salmonella and for this reason their control is exactly the same.

He said that all salmonella are a problem and MDR salmonella is just as much of a problem as non-resistant strains.

He added that MDR salmonella is also no a single serotype, but could be several different serotypes and they do not require special treatment MDR salmonella is not a strain, it is a variety of serotypes and it comes in many different flavours.

Dr Hurd said that studies of slaughter samples of pigs chickens and cattle do not show that there has been a rise in MDR salmonella over the years and he added that the existence of resistance does not necessarily mean that is has been produced by antibiotic use.

"Some salmonellas are naturally resistant to some antibiotics," he said.

"Salmonella are a moving target. They come and they go. They rise and they fall in different populations and different regions."

Dr Hurd said concept that human infections with resistant strains may be harder to treat is a general assumption but humans infected with resistant or non-resistant salmonella will get better at exactly the same rate.

He said that as bacteria has the ability to grow in the presence of some antimicrobials and salmonella has been grown on a plate in a lab in the presence of an antimicrobial, this does not mean that antimicrobials are clinically ineffective and does not mean that treatment is a failure.

Dr Hurd added that all scientific risk assessments have shown that there is negligible risk from antibiotic use on the farm. He said very few people are going to be treated with the antibiotic that are used on the farm when they are treated for a food poisoning event and most doctors will not prescribe antibiotics for food poising in the first place.

He said the research had shown a minimal risk from antibiotics used of the farm and for fluoroquinolone, which is used for heifers on dairy farms the risk is one in 61 billion and for enrofloxacin - a fluoroquinolone - use for poultry the risk to humans was one in 30,000. However, he added that enrofloxacin has been removed from use.

He said that the risk from all antimicrobial uses for cattle, swine and poultry on farm was one in 10 million a year.

"We need to look at the risk and have the data to make decisions," said Dr Hurd.

"But there is a long way from the farm to the hospital. The risk is not zero, but it is very, very low."

Dr Hurd said the greater risk is from the complex link between human health and animal health through some subclinical infection that can cause food borne illness.

He said that if apparently healthy animals with a subclinical illness have some part of the infection left in the carcase at slaughter there is a great possibility that salmonella infection will be passed to humans and the infection can be passed along the slaughterline from carcase to carcase.

He said there is no reason not to use antibiotics on far, particularly as failure to prevent or treat illnesses causes unnecessary suffering and even death to the animals and a lack of treatment increases the illnesses in the animals and increases the risk of foodborne illness in humans.