By Pallab Ghosh
Science correspondent, BBC News
The current system "can over-estimate animal suffering", said the groups
Two bodies that advise the Home Office on animal experimentation have called for a new method for recording the suffering of lab animals.
The Animal Procedures Committee and the Lab Animals Science Association say the suffering of each and every animal used in experiments should be recorded.
At the moment, scientists need provide only a general estimate.
The Royal Society for the protection of Animals, which has been a critic of some lab practices, welcomed the news.
When applying for research licences, scientists have to estimate the number of animals they expect to use and the degree of suffering they will experience, categorised as either mild, moderate or substantial. And it is these figures that are reported in the official statistics released every year.
It is a system that has been criticised by animal welfare groups who argue that it is inaccurate and secretive.
The government's main advisors on the issue agree that an improved system should be introduced. They propose that scientists record the number of animals they use and their degree of suffering as they go along.
According to Professor Dominic Wells of Imperial College London, who helped produce the report, pilot studies of the new system showed that the actual numbers of animals used and the suffering they experienced were much less than scientists had projected.
This resulted, he said, because under the current system scientists provided an estimate of the maximum number of animals they expected to use and also over-estimated their suffering
"I think we'll get a real feel for what happens and at the moment our impression is that actually it will turn out that the vast majority of experimental animals will fall into the mild bracket," he told BBC News.
The proposal has been welcomed by the Royal Society for the Protection of Animals who worked with the APC and the LASA to draw up the proposal.
RSPCA spokesperson Helen Briggs commented: "We think it's fantastic news and we hope these recommendations are something that the government will take forward and actually implement.
"Anything that encourages transparency in these tests is absolutely brilliant news."
The British Union for the Abolition of Vivisection said that although the new system could improve the annual statistics issued by the Home Office, it said that the current "mild", "moderate" and "substantial" bands were "far too vague".
Chief Executive Michelle Thew said: "With no more than 2% of licences categorised as substantial - despite the high number of very unpleasant experiments - this inevitably misleads the public."
Assessments of animal suffering are made by the scientists who carry out the experiments. Ms Thew said that these assessments should be made independently.
"It is clearly in the interests of researchers to downplay suffering for PR reasons and to increase their chance of getting a research licence," she said.
"There has to be a proper system of inspection, properly resourced and truly independent of the animal research industry."
The system proposed by the APC and the LASA will increase the workload on scientists, many of whom complain that there is already too much bureaucracy.
The Home Office said that there should be a public consultation on the new proposals, the results of which would be used to prepare a "regulatory impact assessment".
The number of animals used in UK labs for scientific experiments is currently just over three million a year.