26 October 2001

Changes needed to avoid further research errors

By Richard Allison

CONFIDENCE in animal research has been dented after the mix-up of cattle and sheep brain samples in sheep BSE experiments and industry experts say procedures must change to avoid such a mistake in future.

Flawed projects are now less likely to occur since all research proposals are reviewed independently by a panel of scientists before being funded by DEFRA, explains the Royal Vet Colleges Mac Johnston.

"Back in the 1990s, when many BSE projects were awarded to research organisations, these reviews were not undertaken and the rush to undertake projects in this area may have compromised research quality," he says.

When representing producers interests on BSE, the NFU uses advice from the governments BSE advisory committee, SEAC, says animal health chairman, Neil Cutler.

The union expects all research undertaken by the Vet Lab Agency and the Institute of Animal Health, both implicated in the BSE mix-up, to be reliable, he adds. "Therefore, information from these studies for our policy decisions is used in good faith, particularly from institutes with a good track record in research."

The NFU was preparing to interpret the results of the BSE project this week, says Mr Cutler. "But scientific integrity was never going to be questioned, particularly whether the correct brain samples were tested."

After the sheep BSE study mix-up, the NFU will reconsider its approach to examining important research findings at an animal health and welfare committee meeting this week. "But how do you police scientists? Unless you are an expert you have to take their word," says Mr Cutler.

Prof Johnston questions whether this is an isolated incident and suggests other projects may need to be reviewed. "How valid is other research on which important government decisions have been based?" he asks.

In response to the mix-up, DEFRA has announced an independent audit to find where the mistake was made. Perhaps a similar procedure should be undertaken routinely on all research projects of importance, suggests Mr Cutler.

A separate cross-checking project was done to ensure material was not contaminated with cattle brains, says DEFRA. It was this measure which showed researchers were looking at cattle brains instead of sheep. "But surely any check on sample identity should have been undertaken early in the project," says Prof Johnston.

Mr Cutler agrees organisations such as the Biotechnology and Biological Sciences Research Council, which funds the IAH, should ensure there are adequate procedures to prevent simple mistakes.

But a spokesperson from BBSRC defends its procedures. "All BBSRC funded institutes have been reviewed in the last six months by visiting groups. They achieved a research grading between five and five star, the top score."

These visiting groups consist of researchers, academics, industry representatives and lay people. The group interviews project leaders and considers everything from sample labelling to the quality of science, says the spokesperson.

But there are a number of key stages in a project where checks are essential, says Prof Johnston. This includes adequate sample identification, full sample traceability, reliable lab tests and analysing data using the correct statistical method.

With projects investigating sensitive subjects, the consequences of getting any steps wrong are immense, according to Prof Johnston. "For example, when testing for campylobacter in food, a simple mistake such as allowing samples to dry out can lead to a false result. When one step is in doubt, the whole project becomes flawed.

"But there are several accreditation schemes to ensure high standards in labs. An example is the notifiable horse disease contagious equine metritis, where testing labs are checked annually and those that fail can no longer perform this testing." &#42

ANIMALRESEARCH

&#8226 Must be reliable.

&#8226 Existing checks adequate?

&#8226 Proposals now reviewed.