DNA super-network increases risk of mix-ups

Opening up forensic databases across Europe will make it easier for countries to share data, but it may also lead to miscarriages of justice

PETER HAMKIN was pulling pints in a bar in Merseyside, UK, in 2003, when he was arrested on suspicion of murdering a woman in Italy a year earlier. Italian police had requested a search of the UK DNA database and claimed he was a perfect match, and that he fitted witness descriptions of the murderer. After a 20-day ordeal, a second DNA test ruled Hamkin out and he was released without charge.

Human rights groups and scientists are concerned that mix-ups like this may happen more often once the Prüm Treaty to create a super-network of European DNA databases is implemented.

Interpol already facilitates limited exchange of forensic data between European countries. The new treaty is intended to simplify and speed up this process by making the exchange of information on DNA profiles, fingerprints and vehicle licences more routine.

A number of countries, including the UK, have missed the 26 August deadline for linking up their databases. But it may only be a matter of time before the remaining 16 members of the European Union join the 11 that are already exchanging data.

The worry is that miscarriages of justice will occur without better safeguards in place to reduce the chances of false matches. A key concern revolves around the minimum number of locations on a chromosome, or loci, that police will compare when conducting a search. Labs in the US typically look at 13 loci, while UK labs tend to look at 10. The more loci that are analysed, the more comprehensive the DNA profile will be.

However, the minimum number of loci required to conduct a search under the Prüm Treaty is just six, because European countries often use different loci and there is only limited overlap between them.

Kees Van der Beek of the Netherlands Forensic Institute has calculated the number of false matches that arose when investigators from the Netherlands conducted searches of the German database based on 6 or 7 loci. Of 86 matches generated from 6-locus searches, 57 later turned out to be false positives. Of 276 matches generated from 7-locus searches, 15 were false positives. "Six-locus and seven-locus matches have a real chance of being false positives," says Van der Beek. He says that additional testing of more loci safeguards against false arrests and this does takes place. But if international searches become commonplace, repeat testing may become unfeasible. It is sometimes impossible to retest genetic material that has come from a crime scene and is heavily degraded.

To counter this problem, the European Council has recommended expanding the number of shared European loci by five, but this resolution must be adopted by member states for it to have any effect.

Countries with established DNA databases may be reluctant to incorporate additional loci. "For countries which already have established technical systems, it will be very costly to change," says Eric Toepfer, a consultant for Statewatch, a non-profit organisation that monitors civil liberties in Europe. It would also be impossible to retest all existing samples in the databases.

Either way, more transparency is needed to ensure the justice system works as it should once the database is up and running, says Helen Wallace of Genewatch, a UK-based non-profit organisation. "From the Peter Hamkin case, it's clear that you can have arrests which are based purely on a false match," she says. "The whole area of safeguards around the extradition process and the use of this information in court is critical."