And in England, bartender Peter Hamkin, 23, spent 20 days fearing deportation to face murder charges in Italy in 2002 before Interpol discovered that it had blundered in matching his DNA profile to that of a wanted killer in Tuscany. "I was a prisoner in my own home, constantly on the edge thinking the Italian police were going to arrive to take me away," he said afterward.
In a recent publication of the National Association of Criminal Defence Lawyers, lawyer William Thompson deplored the rash of faked and mistaken DNA results.
"Police and prosecutors have demanded DNA tests in an ever-expanding number of cases, putting pressure on labs to keep pace," he said.
In Canada, the RCMP, Ontario's Centre of Forensic Sciences and a handful of private labs analyze crime-scene DNA.
While CFS has raised its standards since 1995, when an inquiry into the wrongful murder conviction of Guy Paul Morin heard that clothing fibres used to convict him had actually been shed by a lab technician's sweater, an incident last summer was a sobering reminder.
With his first-degree murder trial just two weeks away, Michael Smith of St. Thomas, Ont., learned from the Centre of Forensic Sciences that a defective lot of semen-testing kits several months earlier had yielded a false positive result in his case. "I was shocked that it was a mistake, but also that it had taken so long for them to tell us," said Mr. Smith's lawyer, Robert Upsdell.
Mr. Upsdell asked a top Canadian DNA analyst, John Waye - head of the Molecular Diagnostic Genetics Service of the Hamilton Regional Laboratory Medicine Program - to look into the problem. Dr. Waye recalled that "many, many cases" were affected by the faulty testing kits, including some that had already gone to trial.
While CFS was retesting samples in the Smith case, it discovered that the locations on the victim's body where swabs were obtained had been mislabeled when CFS received them in 2007. "Had the retesting not been required, this mistake would never have been discovered," Mr. Upsdell said.
Dr. Waye said Canadian labs are inspected annually by independent auditors and have improved quality control and tracking of sample. Still, tens of thousands of samples flow in and out every year.
"They do a mind-boggling amount of work, and it is really done as an assembly line. You can do your own job absolutely perfectly, but if somebody messes something up before it gets to you, everything you do is messed up, too."
Leo Adler, a Toronto lawyer who specializes in DNA cases, said most lawyers simply don't believe they can beat a DNA case and seek a plea bargain. "I would say that 99 times out of a hundred, nobody is fighting it any more," he said. "I suspect that there are lawyers whose clients say: 'Hey, it wasn't me,' and the lawyer says: 'You're going to go down, so let's work something out.' "
Mr. Adler said the problem with capitulation is that many cases involve experts making subjective conclusions about DNA sources that were mixed together and in different locations at a crime scene. Technicians can be tempted to stretch their conclusions in the belief that they are helping to convict a dangerous criminal, he added: "When a submission goes in to CFS, it goes along with a police theory."
Ricardo Federico, a Toronto lawyer who issues a monthly DNA newsletter, said he was jolted last summer when an Israeli company, Nucleix Ltd., announced it could synthesize DNA. The news raised the spectre of criminals framing innocent people starting with a couple of microscopic cells.
"Sharp, brilliant minds are not always on the side of law and order," Mr. Adler said, citing a Saskatchewan man who surgically planted a vial of somebody else's blood in his forearm in an attempt to foil a blood test several years ago "People are always trying to stay one step ahead."
Another scientific development that has caused concern is the discovery of individuals who have two distinct DNA strands in their bodies. Known as chimeras, they have unusual DNA profiles that can come about either because of a blood transfusion or because two embryos merged in their mother's uterus. Estimates of the number of chimeras range from a tiny proportion of the population up to 10 per cent.