DNA tests in doubt as court cases collapse

GRAVE DOUBTS hang over the use of genetic 'fingerprinting' in criminal cases - as well as the safety of hundreds of past convictions - after an Old Bailey judge questioned the reliability of methods used by Metropolitan Police scientists.

The ruling, by Judge Alan Rawley, could pave the way for a flood of appeals in murder and rape cases where convictions were obtained solely or partly through evidence based on DNA testing, which has been presented to courts as a fool-proof scientific method.

The courts first accepted DNA evidence in 1987. It works by comparing samples of human tissue found at the scenes of crimes - hair, semen or blood, for example - with tissue samples taken from suspects. Convictions have been gained after prosecutors told juries that there was only a one in a thousand or even one in several million chance that the DNA evidence 'proving' the defendant guilty could be the result of a simple coincidence.

Judge Rawley's judgment came after lawyers representing Terrence Hammond, 28, accused of armed robbery, said the procedure used by police forensic scientists was unsafe. Given two conflicting sets of 'expert' witnesses, Judge Rawley ruled the DNA evidence inadmissible, forcing the Crown Prosecution Service to drop all charges.

His decision follows a similar ruling by an Old Bailey judge in a rape case last month and will prompt prosecution barristers to question whether they can present such evidence in future trials.

Doubts about the use of DNA tests arose in the US two years ago when a group of eminent geneticists argued that statistics used to assess the probability of a chance match between suspect and forensic samples were flawed.

Similar arguments were used in the Hammond trial. Thomas Fedor, a forensic scientist who once worked for the Detroit police, was called as a defence witness. He alleged that:

The Metropolitan Police and Home Office used different guidelines to agree a match between crime sample and the suspect's blood;

The probabilities used to assess a chance match were arbitrary and based on a reference sample of DNA fingerprints that was not representative of the population;

A statistical method employed by the police of multiplying probabilities to reach fantastically high chances of a coincidental match should not be used;

The police used a statistical procedure devised by Home Office scientists that has not been published and was not open to scrutiny by independent experts.

In the Hammond case, the prosecution initially said the probability of a chance match between his DNA sample and the blood at the scene of the crime was about 350 million to one. This was later rounded down to 10 million to one for safety.

However, Mr Fedor argued that such figures could not be supported. He said the Detroit police has suspended the use of DNA tests because of 'uncertainties in the US as to whether these numbers are reliable or not'.

Last night, Scotland Yard and the Crown Prosecution Service said they intended to continue presenting DNA evidence in court.