The Guardian April 19, 2000

Genetic fanatics

This is part one of an abridged version of the "briefing paper" presented 
to the NSW Parliament on April 6 by Justice Action. Part two will appear in 
a future issue of The Guardian. The full document can be obtained 
from Justice Action, PO Box 386 Broadway, NSW 2007; Ph: (02) 9281 5100 Fax: 
(02) 9281 5303;


* * *
There has been much recent enthusiasm expressed about the potential of forensic DNA testing to solve crime and even to free the wrongly convicted. This enthusiasm has led to a rush to implement new technology without considering possible drawbacks or safeguards. It has also led to proposals which would severely compromise privacy and civil rights were they to be implemented in NSW. The Technology The public has been regaled with stories of the new wonder technology which will solve crime using a single hair or speck of saliva, that will positively identify the criminal with one-million-to-one certainty, freeing the innocent, condemning the guilty and ushering in a brave new world of crime fighting. The truth is somewhat less spectacular. The older technology currently used (Restricted Fragment Length Polymorphism or RFLP) is subject to errors in each of its six stages. The newer technology (Polymerase Chain Reaction amplification or PCR) requires better quality, though smaller, samples and extremely high lab standards to be successful. Due to cross contamination between samples from a Christchurch assault scene and two Wellington murder scenes, a NZ man spent four months under police investigation for double murder. Fortunately he had an airtight alibi: he was in Christchurch being assaulted. NZ Justice Minister Phil Goff said of the case: "What worried me was that, okay, this was a very clear-cut case where the man couldn't have committed the crimes. Had he been a gang member and lived in Wellington ..." It needs to be remembered that DNA testing does not produce "fingerprint" like results positively linking a suspect with a crime scene, rather it produces a statistical probability such as "there is a one in a million chance that a randomly selected person will have a DNA profile as close to the sample as that of the suspect". If all of NSW's 8000 prisoners were tested against the 15,000 profiles currently in the NSW police database simple chance indicates that around 120 false positives would result. Should such a test be applied to the Australian population as a whole, the labs would find around 300,000 "culprits" for the 15,000 unsolved crimes. This has serious implications when tests are performed against an accused member of an ethnic minority. While the chance that a randomly selected Australian citizen would have a DNA profile matching a Brewarrina Aborigine may be a million to one, a randomly selected Brewarrina Aborigine may be at a much higher chance of matching. Some enthusiasm has been expressed, by NY police chief Howard Safir among others, for the potential of mitochondrial DNA from hair to be used to profile even the most degraded samples. While mitochondrial DNA might be used to aid in the identification of corpses by linking them with female ancestors, it is identical in all individuals of the same female line of descent. Therefore a single profile is common across large sections of the population, even those who do not share a common surname or known family relationship, making it entirely unsuitable for cases of criminal identification. The probabilities quoted in court regarding DNA evidence completely neglect the error rates of the testing laboratories. An external DNA proficiency test conducted by the California Association of Crime Laboratory Directors upon the Lifecodes, Cellmark and FBI laboratories found a one percent error rate, mostly false positives. Odds like "one in a million" are meaningless when the lab error rate is one in a hundred. In the UK, forensic DNA testing laboratories are subject to random blind testing to enable their error rates to be accurately estimated. Procedures in US FBI laboratories were condemned following a 1997 Justice Department investigation which revealed deliberate faking of data to obtain convictions. While the new technology may look impressive, ways of rorting it are equally impressive. It requires no expertise in fingerprint lifting and transfer to plant genetic material on a crime scene and it is far less risky than "loading someone up"with heroin. In the US several innovative methods are already in use to foil genetic testing, including prisoners spitting into each other's mouths before a swab and a case where an imprisoned rapist sent a sample of his semen to a female friend so that she might register a rape case with a DNA profile exactly matching his. As the technology becomes more common, doubtless so too will ways of beating it. Data and samples The genetic material contained in a forensic sample holds far more intimate information about the subject than that contained in a fingerprint. Physical, racial and behavioural information about the subject will all be available to the skilled geneticist of the near future, with the completion of the Human Genome Project. The data available from these samples can be expected to become more and more valuable as genetic science progresses. This information will be of obvious use to drug and insurance companies but the coming decades will see new and unimagined commercial applications for such data. The data extracted from these samples with current forensic technology is far more limited, but still contains information that may be used other than in criminal investigation. The existence of the samples in storage or the data on computers will provide unique temptations to corruption for those charged with accessing them and will present unique security problems due to the data value of a microscopic speck of material. If this data is to be stored on computers with internet connections it can be expected to become the target of attacks from well resourced hackers with an interest in duplicating or altering it. It is important that all police and lab staff with access to such material and data be made aware of its sensitivity and the need to ensure its security. All samples should be immediately destroyed when they are no longer required for the case being investigated or tried. PCR technology involves the mass duplication of part of a subject's DNA. In this age of "patenting" DNA sequences it is important that the donor retain ownership of all information contained in his/her DNA, regardless of how it may be copied or disseminated. Following testing, DNA laboratories should be required to surrender or destroy any material which may contain originals or copies of a subject's DNA as well as any digital or physical record of such data. The genetic profiles resulting from the tests should only be entered onto a police database following a conviction for a serious offence. Fingerprint records are destroyed if a conviction is not obtained although they do not contain information which might be used for other purposes and provide positive identification, unlike DNA tests. The police should not have access to a database of test results from innocent people to allow them to "go fishing" for matches to unsolved crimes. As discussed above, the technology is completely unsuitable to such mass cross-matching and is no substitute for traditional detective work. Samples and data from crime scenes which remain unlinked to a specific individual require no such protection however, and represent a valuable resource both in the solving of crime and the possible acquittal of the wrongly convicted. It is vital that defendants and convicts be given full access to crime scene DNA samples where these samples may establish their innocence. Procedures in many US States make it very difficult to gain such access and of the 70 convicts freed when DNA evidence overturned their conviction, more than half had to take legal action to gain access to the samples, often signing away their right to compensation in the process. Sample collection Although its enthusiasts refer to DNA profiling as "genetic fingerprinting" it is both less precise and far more invasive than traditional fingerprinting. Modern techniques allow fingerprints to be taken without even the inconvenience of an inky hand. DNA sampling requires the insertion of a needle into the flesh or a swab into a bodily orifice. Legislation in Victoria allows for prisoners to be held down and a swab to be forcibly inserted into their mouths. This is the closest thing to legalised rape since cavity searches for drugs. Yet body searches and fingerprinting can only be done when someone is strongly suspected of a criminal act. Under current proposals all NSW prisoners will be subject to such treatment in spite of being under no suspicion of committing an unsolved crime. There is also a proposal for all those arrested for any offence in NSW to be subject to such an invasion of their body. Even suspicion of serious criminal activity does not of itself justify DNA testing. Only in a tiny minority of cases is DNA evidence relevent to investigation or prosecution.

Back to index page